Conference Agenda

We report on the main results of a recent work [1], in which we describe the reorientation of reinforcing fibers in a biological medium by blending together concepts of continuum mechanics and statistical mechanics.

Assuming saturation, fiber-reinforced tissues –such as articular cartilage– are often described macroscopically as triphasic media consisting of an interstitial fluid, a solid matrix and the fiber phase. Typically, the fibers are oriented non-uniformly in space and are assumed to be embedded in the matrix, which can be viewed as a double-porosity medium [2]. Moreover, while the matrix is often hypothesized to be isotropic, the complex matrix-fibers is generally anisotropic due to the orientation and material properties of the fibers. This space-dependent anisotropy also impacts the fluid flow by intervening, for example, on permeability.

In the literature (see, e.g., [3–6]), the orientational pattern of the fibers is often described by means of a probability density distribution, defined as a function of a field of unit vectors [7], and employed to determine the directional averages of the constitutive functions of interest written for a generic spatial direction.

Usually, the functional form of the probability density distribution is interpreted as a known property, assigned from the outset. This applies also to some models of fiber reorientation in which the probability density distribution maintains the same dependence on the field of unit vectors, although being parameterized by variables evolving in time [8,9]. Such variability, attributed for instance to the direction of the fibers’ most probable orientation, describes a type of tissue structural reorganization referred to as remodeling.

In our work, we deviate from the models in which the probability density distribution is assigned a priori and, by assuming that the fiber dynamics is governed by a Fokker–Planck equation, we compute it as the stationary solution of this equation. Specifically, we hypothesize that, at the fiber scale, reorientation is a dynamic process of Langevin type driven by a deterministic contribution and a stochastic one: the former is the deformation of the matrix; the latter resolves the short-range reciprocal actions among the fibers. By following this approach, we are able to retrieve relevant probability density distributions, such as the von Mises one, as particular stationary solutions of the Fokker–Planck equation. We also account for other structural transformations occurring in the matrix by assuming that they are represented by inelastic distortions.

References

[1] Giammarini, A., Pastore, A., Grillo, A.: Math. Mech. Solids, In production.

[2] Tomic, A., Grillo, A., Federico, S.: IMA J. Appl. Math., 79: 1027–1059 (2014).

[3] Federico, S., Gasser, T.: J. Roy. Soc. Interface, 7 955–966 (2010).

[4] Holzapfel, G.A., et al.: J. Roy. Soc. Interface, 12(106) 20150188 (2015).

[5] Gizzi, A., Pandolfi, A., Vasta, M.: J. Eng. Math., 109(1) 211–226 (2017).

[6] Holzapfel, G.A., Ogden, R.W.: J. Elasticity, 129(1–2) 49–68 (2017).

[7] Lanir, Y.: J. Biomech., 16 1–12 (1983).

[8] Baaijens, F., et al.: J. Biomech., 43 166–175 (2010).

[9] Grillo, A., et al.: J. Eng. Math., 109(1) 139–172 (2018).

Skeletal muscles are a biological tissue with unique mechanical properties, characterised by anisotropic effects determined by its fiber-like microstructure. It features a nonlinearly elastic response and, most importantly, its physiological motion is determined by activation phenomena that need to be included in the continuum models for its behaviour. Of particular relevance is also the multiscale nature of the muscle tissue. Indeed, sarcomeres, the smallest contractile units, are bundled to form miofibrils, which in turn constitute the muscle in its large-scale configuration, with non-trivial geometric arrangements. Experimental evidence indicates that the mechanical response of the smaller units is rather different from and simpler than what can be measured for the whole muscle. Hence, a multiscale approach is necessary to arrive at both an understanding and a proper continuum modelling of muscle dynamics.
We first present a theoretical framework for anisotropic nonlinear elasticity based on the decomposition of strain and stress tensors on a tensorial basis adapted to the local anisotropy of the material. The definition of a local orthogonal basis for the space of second-order symmetric tensors allows to express a generic constitutive prescription as a vector field on a six-dimensional space, in which each of the six components represents an independent and objective material function. The presence of local anisotropies is reflected on material symmetries, and we consider the corresponding restrictions on material functions for some important crystal classes and for transversely isotropic and isotropic materials. This formalism aims at a clear and mechanically motivated organisation of the degrees of freedom involved in describing nonlinear elasticity, to facilitate the experimental identification of material functions for their constitutive characterisation.
Within the said framework, we start from experimental data about the passive response of muscle specimens to identify nonlinear material functions for the stress-strain relation, thereby following a data-driven approach. The peculiar nonlinearity of the response can be ascribed to the microscopic heterogeneity that leads to a progressive recruitment of different fibers, ultimately leading to the strain-hardening evidenced by the data. A careful analysis of this phenomenon, driven in particular by its anisotropic nature, gives also important information about the asymmetry of the material response under traction or compression. We propose a basic model in the context of anisotropic Cauchy elasticity to capture these phenomena and then propose a strategy to include an activation mechanism. Finally, it is important to consider the significant sources of dissipation that affect the muscular dynamics. For this reason, we explore how to include in the constitutive law terms that produce a viscus damping during the passive and active motion.

The structural and refractive capabilities of the eye depend on the mechanical properties and imperfections that are present in its constituents, e.g. cornea, aqueous and vitreous humor. In particular, the cornea is a hydrated biological tissue with the structural role of balancing the mechanical loading exerted by the physiological intraocular pressure, and it is characterized by an underlying collagen microstructure [1]. Moreover, the cornea can be classified as a multi-constituent material, since it is possible to distinguish at least a solid and a fluid phase. In particular, the solid phase comprises various constituents as keratocytes, collagen fibers and extracellular matrix, while the fluid phase is constituted primarily by water and carries out electrostatic charges, which circulate due to passive transport and active pumping action of the corneal endothelium. For instance, these ionic charges contribute to the osmotic pressure experienced by the tissue [2].

In our work, by adhering to a thermodynamically consistent framework [3], we describe the cornea as a poroelastic medium undergoing finite deformation, and we show some numerical simulations within physiological parameters for a comparison with the results already present in the literature. Moreover, we are interested in investigating the predictive capabilities of our model when describing the pathological condition known as keratoconus, in which, due to a mechanical failure, the cornea assumes a conical shape [4]. While the cause of keratoconus insurgence is currently unclear, previous studies linked the change in shape of the cornea with an alteration of the fiber microstructure, and with a degradation of the elastic properties of the tissue [5].

In this respect, our aim is to improve the mechanical description of the progress of the disorder by introducing the effects related to the circulation of the fluid. The degradation of the mechanical properties of the cornea is accounted by introducing a damage variable, as is done in the literature for the monophasic models [5]. The introduction of the fluid phase requires the prescription of appropriate boundary conditions, and we perform finite element numerical simulations to assess qualitative and quantitative results. We remark that, to the best of our knowledge, this is the first attempt at formulating a poroelastic model of the cornea that accounts for the fluid’s mechanical influence, and it represents a first step towards a complete model that incorporates both the mechanical and hydraulic role of the endothelium.

References

[1] Meek K. M. et al. Current Eye Research, Vol. 6, No. 7, Informa UK Limited, p. 841-846 (1987).

[2] Lewis N. P. et al., Structure, Vol. 18, No. 2, Elsevier BV, p. 239-245 (2010).

[3] Hassanizadeh S. M., Advances in Water Resources, Vol. 9, p. 207-222 (1986).

[4] Rabinowitz Y. S., Survey of Ophthalmology, Vol. 42, No. 4, Elsevier BV, p. 297-319 (1998).

[5] Pandolfi, A., Mechanics of Materials, Vol. 199, Elsevier BV (2024).

Introduction

Tumour growth is a complex mechanosensitive process guided by feedback between cells and the extracellular matrix (ECM). However, the underlying biomechanisms by which external loading impacts tumour growth remain unknown. Here we develop an in-vitro culture system for heterogeneous tumour spheroids to identify key biomechanical factors that restrict spheroid growth. We further propose a novel framework consisting a novel hydromechanical cell growth model, a 3D deformable cell framework and a deep-neural network (DNN)-accelerated finite element (FE) solver to uncover the mechanisms underlying mechanosensitive growth.

Methods

Murine breast cancer (4T1) cells were cultured, isolated and propagated to obtain tumour spheroids of distinct phenotypes within gelatin hydrogels of varying stiffness to investigate stress-dependent responses. A mathematical model was developed to predict cell growth as driven by a competition between hydrostatic pressure arising from active cell stress and external loading, and osmotic pressure arising from biomolecule synthesis and ion fluxes [1]. Biomolecules are synthesized during the G1 growth phase, which drives a fluid influx and growth through entropic osmotic forces. This model was integrated with MPacts [2], a discrete element method platform for multicellular mechanics. Loading from cell and matrix contact interfaces governs the cell growth rate, and we consider that mitosis is subject to surpassing a critical volume checkpoint for which the underlying mechanisms are characterised. The deformable cell model is embedded in a 3D finite element model of a hyperelastic matrix. To efficiently simulate this matrix deformation, a deep neural network (DNN) framework was developed. Synthetic training data for a diverse range of loading scenarios was generated and subsequently provided to a fully connected DNN to obtain a converged solution.

Results and Discussion

Analysis of the in-vitro spheroids reveals that tumour spheroid size reduces with increasing gel stiffness. Importantly, there was a significant reduction in the number of cells per spheroid and lower number of proliferative cells as determined by Ki-67 immunofluorescence. 4T1 spheroids developed from predominantly amoeboid-like cells were observed to have 2-3 fold larger diameters than epithelial-like spheroids. Our computational models predict that cell growth increases with biomolecule synthesis and the subsequent increase in osmotic pressure, which deforms surrounding matrix. Tumour spheroid growth simulations suggest that cells at the tumour core experience higher pressure than cells at the periphery, as previously shown experimentally [3]. These cells are also predicted to exhibit higher levels of biomolecule crowding. Our models indicate that cell confinement and associated loading leads to an imbalance between hydrostatic and osmotic pressures, which drives fluid loss and ultimately a reduction in cell growth below the volume checkpoint and suppression of proliferation. Importantly, our simulations predict a reduction in spheroid size with increasing matrix stiffness in agreement with our experimental data, and correctly capture cell morphologies and spheroid compaction. The fully coupled multicellular model allows for a deep insight into the mechanics of tumour growth at the cellular level.

References

1. McEvoy, E. et al. (2020). Nat. Commun.

2. Ongenae, S. et al. (2024). biorXiv

3. Nia, H. et al. (2017). Nat Biomed Eng

Traditionally, cancer has been viewed as a disease originating from a single cell that accumulates genetic mutations, eventually leading to tumour formation. However, recent research emphasizes the crucial role of the tumour microenvironment, especially its mechanical properties, in influencing cancer progression. Understanding the tumour microenvironment is key to gaining deeper insights into tumour development and response to therapies.

To investigate these complex interactions, researchers rely on both in vitro and in vivo models. In vitro models provide controlled environments to study specific tumour behaviours, while in vivo models offer more realistic biological contexts. However, experimental methods are often limited by their inability to isolate specific factors and efficiently test new hypotheses.

Computational modelling provides a powerful complementary approach. These models can mimic experimental conditions, test novel ideas, and explore system characteristics that are otherwise difficult or impractical to examine experimentally. Nevertheless, for computational models to be effective, they must be carefully calibrated using experimental data and account for both computational and experimental uncertainties.

In this work, we develop a comprehensive workflow that enables the integration between experimental data and computational tumour models. The process begins with the formulation of a flexible multiphase poroelastic model capable of simulating tumour growth, where the extracellular matrix is represented as a solid scaffold and the tumour and healthy cells, along with interstitial fluid, occupy the pore space. Once the computational model is established, we perform a global sensitivity analysis on a reduced-order version to identify the most influential parameters driving tumour behaviour. This step not only helps to simplify the model but also guides the design of future experiments by highlighting key variables. Building on this, we apply Bayesian inference techniques to explore and estimate the values of these parameters, accounting for uncertainty and variability inherent to biological systems. By treating parameters as probability distributions rather than fixed values, the model achieves a more realistic and robust representation of tumour dynamics. The proposed workflow is successfully applied to a case study of tumour spheroid growth, demonstrating its potential to bridge experimental observations with predictive computational modelling in a systematic and efficient way.

The vertebrate limb has long served as a paradigm to understand morphogenesis. Numerical models have played a key role in understanding how cells self-organize based on positional cues and local interactions to specify their fate, thereby driving skeletal patterning. Classical approaches typically rely on reaction-diffusion systems and positional information frameworks, and recent work has begun to explore their integration [1]. While these approaches have yielded valuable insights into localized patterning, they often fail to explain how multiple structures emerge sequentially and coherently within a growing domain.

In this work, we introduce a computational framework that integrates growth dynamics with self-organization and positional cues to predict the full skeletal pattern of a vertebrate limb for the first time. The new Growth-Processing-Propagation (GPP) framework [2] is formulated as a system of reaction-diffusion equations on a time-evolving domain. The model is governed by two non-dimensional parameters, α_R and β_D, which represent the ratio of growth rate to reaction and diffusion, respectively. These parameters determine the influence of bio-signal processing (reaction) and communication (diffusion) in relation to the growth rate on skeletal pattern formation. Unlike traditional models that rely on static or idealized domains, GPP captures the dynamic interplay between limb growth and patterning mechanisms in both space and time.

We apply the GPP framework to axolotl limb development, a classical system for studying morphogenesis due to its accessibility and regenerative capabilities. To constrain the model, we use time-lapse imaging to extract real limb geometries in developing axolotls. Additionally, we incorporate fluorescence images of molecular markers to define positional inputs and to validate the predicted skeletal patterns, focusing on the sequential emergence of the stylopod (upper arm), zeugopod (forearm), and autopod (digits).

The model reproduces the order and spatial distribution of skeletal elements without explicitly specifying when or where each segment should emerge. The number of digits and spacing emerge from a Turing-like instability modulated by domain growth, while transitions between limb segments arise naturally from growth-driven changes in effective patterning timescales. This suggests that variation in growth rate across space and time can shape skeletal complexity by influencing the dynamics of pattern formation.

To our knowledge, this is the first computational model to predict complete skeletal limb patterning, constrained by experimental measurements of domain shape, growth, and molecular marker distribution. The GPP framework offers a generalizable and interpretable modeling approach for simulating pattern formation on growing domains. It bridges classical morphogen-based theories with mechanical expansion, and provides a compact set of parameters to explore biological design space. We anticipate this approach will be broadly applicable to problems in morphogenesis, developmental and regenerative biology, where understanding the coupling between growth and patterning is essential.

[1] Raspopovic et al. (2014), Science 345(6196):566-70, doi: 10.1126/science.1252960.

[2] Ben Tahar et al. (2025), bioRxiv preprint, doi:10.1101/2025.03.20.644440.

Biological systems leverage gradients to direct structural organization and assembly with remarkable precision. In contrast, achieving precise structural control in synthetic materials remains challenging. Here, we report a gradient-mediated self-assembly mechanism where insect cuticle peptides (ICPs) spontaneously form nanocapsules through a single-step solvent exchange process. The concentration gradient formed by water-acetone diffusion directs peptide localization and self-organization, enabling precise structural control without templating. Molecular dynamics simulations reveal how solvent interactions modulate peptide binding and assembly. This work provides insight into mechanics-guided self-assembly, offering strategies for biomimetic nanomaterials, drug delivery, and responsive biointerfaces.

Patient-specific cardiac simulations require the calibration of physical model parameters from measurements of the patient's physiology. Parameter estimation in the heart has been conducted so far using measurements in the myocardium, typically displacement surrogates obtained from MRI. To the best of the authors' knowledge, parameter estimation in cardiac mechanics using velocity images of the ventricular blood flow has remained unexplored. This study assesses the use of Eulerian volumetric velocity images of the ventricular blood flow for estimating material properties of a physiological fully-dimensional fluid–structure interaction (FSI) model of the systolic phase of the heart contraction.
The myocardium is modeled as a hyperelastic material using the standard Holzapfel-Ogden constitutive model with an active contribution to the stress, accounting the contraction of the heart. A fractional step scheme is used to simulate the blood flow efficiently, splitting the Navier–Stokes equations governing the fluid into a tentative velocity step and a projection step to be solved for the fluid pressure. Writing the problem in the Arbitrary Lagrange–Eulerian formalism avoids remeshing the deforming fluid domain at every timestep. In the FSI algorithm, the solid and the fluid problems are coupled in a semi-implicit fashion. In particular, the fluid mesh update and the velocity step are carried out explicitly, depending on the solid displacement of the previous timestep. The fluid's projection step is coupled implicitly to the solid problem, which gives rise to a nonlinear problem, solved with a Krylov–Newton method.
The study considers the left and right ventricles extracted from medical images of a patient. Valve dynamics are not included, as the semilunar valves are considered wide open and the atrioventricular valves closed during systole.
Synthetic measurements of both the solid displacement and the fluid velocity are generated by computing a solution of the FSI problem with fixed ground truth parameters, subsampling the solution in time and adding noise at levels typical for medical images.
The cardiac mechanics model parameters estimated from these measurements are (a) the myocardial tissue contractility, controlling the active contraction, and (b) the epicardial stiffness, characterizing the epicardial wall boundary condition by accounting for the external tissue support. Sequential data assimilation, namely a reduced-order unscented Kalman filter, is employed to estimate the parameters from (i) solid displacement measurements only, (ii) fluid velocity measurements only, or (iii) a combination of both.
Our findings indicate that the accuracy of the estimated parameters with respect to the ground truth strongly depends on the temporal resolution of the data, with only high resolution data allowing for accurate parameter estimates. Using combined fluid and solid measurements reduces the sensitivity of the estimation to measurement noise. The contractility estimation result is closest to the ground truth when using fluid measurements only, whereas the epicardial stiffness is recovered more accurately when using only solid measurements.

Accurate estimation of electrophysiological parameters is crucial for understanding and modeling the dynamic behavior of cardiac tissues across the cardiac cycle, which includes phases of depolarization, repolarization, and refractory periods. These phases underpin the generation and propagation of action potentials that coordinate rhythmic contractions of the heart. In this study, we propose a robust parameter estimation framework based on Differential Evolution (DE) to infer hidden parameters of nonlinear cardiac models. DE, a population-based global optimization algorithm, is particularly suited for handling the nonlinearity and high dimensionality inherent in biological systems. We first demonstrate the effectiveness of DE using the Van der Pol oscillator, a simplified excitable system that mimics the cyclical behavior of cardiac cells. Building on this foundation, we apply our method to more detailed and physiologically relevant models, including the FitzHugh-Nagumo model, which captures essential features of excitability and recovery in cardiac membranes, and the O'Hara-Rudy (ORd) model, which represents detailed ion channel kinetics, intracellular calcium handling, and membrane potential dynamics in human ventricular myocytes. These models are critical for simulating normal and pathological conditions such as early afterdepolarizations, arrhythmias, and conduction blocks. Our results show that DE can accurately recover key parameters even in the presence of sparse or noisy measurements. This highlights DE's value as a powerful tool for inverse modeling in computational cardiology, enabling deeper insight into the mechanisms of cardiac excitability, arrhythmogenesis, and potential therapeutic interventions.

In computational biomedicine, model geometry generation is commonly based on images, which are typically assumed to be at rest and free from external stimuli. However, biological structures are often in mechanical equilibrium under various forces or transitioning between states, necessitating a more dynamic modeling approach. We present a framework that extends the classical formulation of stress-free geometry computation, known as the inverse elasticity problem (IEP), to fully nonlinear poroelastic media. Our approach involves expressing the governing equations in terms of the reference porosity and defining a time-dependent problem where the steady-state solution corresponds to the reference porosity. We introduce Anderson acceleration as a means to significantly enhance computational speed, achieving up to an 80% reduction in the number of iterations. Furthermore, we identify an inconsistency in the primal formulation of the nonlinear mass conservation equations, arising from second-order derivatives of the strain, which we resolve using appropriate mixed formulations.

A major challenge in advancing functional organ assessment is the limited understanding of vascular development and morphology. To address this, we present SynGROW (Synthetic Generation of Rigorously and Rapidly Optimized Volumetric Vasculature) – our open-source computational framework for generating and analyzing synthetic vascular trees using rigorous global optimization techniques [1]. Our objective function minimizes both the energy required to maintain the vascular structure and the energy required to pump the blood through it, while incorporating patient-specific information from imaging data. Our physiological model does not impose Murray’s law of optimal branching explicitly; instead, we show that it naturally emerges from the optimization process [2]. Our model also incorporates variable blood viscosity at different levels of the vessel hierarchy due to the Fåhræus-Lindqvist effect.

SynGROW integrates global geometry and topology optimization into a unified algorithm that solves a nonlinear programming (NLP) problem with super-linear efficiency. Unlike existing methods limited to single-tree generation in convex domains, SynGROW contains various algorithmic extensions that enable the generation of multiple non-intersecting vascular tree structures within complex nonconvex anatomical domains [3].

In benchmark tests on brain tissue, SynGROW outperforms state-of-the-art methods based on local algorithms, computing results an order of magnitude faster while significantly reducing the objective function. Its results closely match experimental liver vasculature data, accurately capturing physiological features such as tortuosity and parallel branching. Beyond benchmarks, we demonstrate its integration within functional imaging and multiphysics modeling workflows. In the brain, it predicts spatial patterns of penetrating arteries and veins aligned with observed microcirculation [4]. In the liver, it serves as a foundation for modeling hepatic hyperperfusion in the context of liver regeneration [5].

References

[1] E. Jessen, M.C. Steinbach, C. Debbaut and D. Schillinger. Rigorous mathematical optimization of synthetic hepatic vascular trees. J. R. Soc. Interface 19:20220087, 2022.

[2] E. Jessen, M.C. Steinbach, C. Debbaut and D. Schillinger. Branching exponents of synthetic vascular trees. IEEE Trans. Biomed. Eng. 71:1345-1354, 2024.

[3] E. Jessen, M.C. Steinbach, and D. Schillinger. Optimizing non-Intersecting synthetic vascular trees in nonconvex organs. IEEE Trans. Biomed. Eng. (in press).

[4] L.B. Glandorf, E. Jessen, et al. Deciphering morpho-functional stroke repercussions on the cortical microvasculature with Bessel beam OCM. In: Proc. Optical Coherence Tomography and Coherence Domain Optical Methods in Biomedicine XXIX, SPIE, 2025.

[5] A. Ebrahem, J. Hohl, E. Jessen, M.F.P. ten Eikelder and D. Schillinger. A multiscale-multiphysics framework for modeling organ-scale liver regrowth. J. Mech. Phys. Solids 200:106113, 2025.

The diagnosis of cardiac ischemia in acute conditions is of utmost importance in the clinical practice since several therapeutic decisions could follow. In the first part of this talk, we introduce a new cardiac perfusion model [1], accounting for the coronary compliance and for the diastolic flow of microcirculation, for the non-invasive prediction of blood flow maps at the myocardial level. We also discuss the challenging issue of its calibration to describe personalized conditions of patients. Finally we show an application of this model to a wide cohort of patient coming from Centro Cardiologico Monzino in Milan.

In the second part of the talk, we address the issue of assessing the arrhythmogenic risk, in terms of development of ventricular tachicardia, in presence of acute ischemic regions. To do this, we start from the maps discussed in the first part of the talk and we present a computational model [2] based on the local modification of the ionic currents to account for specific processes (such as hyperkalemia and hypoxia) which are consequences of the presence of acute ischemia. We present results of numerical experiments aiming at evaluating, among the others, the effect of the ischemic heterogeneity on the arrhythmogenic risk.

All the numerical experiments were performed with the Finite Elements library LifeX, developed at MOX, Dipartimento di Matematica, in collaboration with LaBS, Dipartimento di Chimica, Materiali e Ingegneria Chimica, both at Politecnico di Milano

REFERENCES

[1] Montino Pelagi G., Regazzoni F., Huyghe J.M., Baggiano A., Ali' M., Bertoluzza S., Valbusa G., Pontone G., Vergara C., Modeling cardiac microcirculation for the simulation of coronary flow and 3D myocardial perfusion. Biomech. Model. Mechanobiol., 23, 1863-1888, 2024.

[2] Corda A., Pagani S., Vergara C., Influence of acute myocardial ischemia on arrhythmogenesis: a computational study, medRXiv, DOI:10.1101/2024.11.20.24317476, 2024

Introduction: Aortic stenosis (AS) is the most prevalent primary valve disease requiring intervention. Transcatheter aortic valve implantation (TAVI) has become the preferred treatment for elderly patients at high surgical risk. However, procedural complications remain a major concern. Among these, conduction abnormalities—affecting 2% to 34% of cases—are primarily caused by excessive mechanical stress exerted by the bioprosthesis on the membranous septum, often leading to the need for permanent pacemaker (PPM) implantation. Additionally, paravalvular leakage (PVL), observed in 7% to 40% of TAVI patients, primarily results from an inadequate sealing between the prosthesis and the aortic annulus.

To address these challenges, in-silico technologies are increasingly used to simulate TAVI procedures. However, existing models have limitations in predicting complications, and their high processing time hinder their clinical applicability. For this reason, this study aims to develop a machine learning (ML) model trained on clinical data and to compare its predictions with the outcomes of patient-specific TAVI simulations. By integrating these approaches, we seek to create a comprehensive predictive profile for each patient, enabling the prognosis of conduction disturbances and PVL. Ultimately, this could potentially replace complex numerical simulations, facilitating more efficient treatment planning.

Methods: A total of 1,897 patients undergoing TAVI were enrolled and their clinical, laboratory, echocardiographic, and procedural data were collected. After conducting exploratory data analysis and feature engineering, dimensionality reduction was performed with UMAP. A Topological Data Analysis (TDA) Kaplan Mapper was then applied to identify patient phenotypes more susceptible to complications. To address class imbalance, undersampling of the majority class was performed, ensuring a 1:1 ratio. After that, various machine learning classifiers were trained and evaluated for their ability to predict conduction abnormalities and PVL.

Concurrently, an external cohort of 50 patients was recruited for in-silico modeling. TAVI simulations were conducted, incorporating patient-specific anatomy, the implanted bioprosthesis, and arterial pre-stress. TAVI digital twins were then validated against post-procedural clinical data. Finally, the external patient cohort will serve as an independent validation set for the ML model. Predictive outcomes will be analyzed alongside patient-specific simulations to assess the model’s limitations and improve its interpretability.

Results: TDA revealed effective clustering of patients who developed moderate or more severe PVL and required PPM implantation. The ML models demonstrated strong performance, with an accuracy of 0.79 and an F1 score of 0.80 for predicting PVL, and an accuracy of 0.76 and an F1 score of 0.78 for predicting PPM implantation. The classifiers that yielded the best performance were XGBoost and Random Forest, respectively. In terms of in-silico modeling, TAVI digital twins successfully replicated valve positioning within the aortic root, as validated by angiographic images and post-operative CT stent segmentation, with errors lower than 5% in terms of orifice area. Furthermore, diastolic CFD simulations were successfully completed.

Conclusions: Our findings are highly promising, demonstrating effective AI-based prediction of complications and accurate in-silico reproduction of the TAVI procedure. Refining ML models and integrating them with patient-specific simulations could significantly enhance clinical practice by providing a robust, rapid, and automated tool for pre-operative planning.

Cardiac radiofrequency ablation (RFA), a cornerstone treatment for arrhythmias, suffers from limited understanding of lesion formation, hindering optimal outcomes. Current RFA models often oversimplify the complex multiphysics interactions within heterogeneous domains, neglecting crucial factors. This research introduces a high-fidelity, multiphysics, and multi-domain computational framework designed to enhance RFA treatment and minimize complications. The framework integrates heat transfer, electrostatics, fluid dynamics, and a three-state cell-death model across electrode, fluid, and tissue regions. Notably, some processes are confined to specific compartments (e.g., fluid dynamics in the fluid, cell-death in the tissue), while others (e.g., electrostatics, heat transfer) span multiple domains. To achieve accurate and scalable simulations, we employ physics- and domain-decomposition (DD) approaches. Specifically, Dirichlet-Neumann and Optimized Schwarz-like DD methods are explored, supported by rigorous convergence analysis. Implemented using high-order numerical methods within the MFEM library, and enhanced by efficient partially assembled operators and ongoing GPU acceleration, this framework demonstrates significant computational efficiency. Importantly, the framework's design allows for extensibility, making it a versatile platform for RF simulations across diverse tissue types (e.g., kidney, uterine, hepatic) and energy sources (RF, microwave, ultrasound, laser), as well as a range of other biomedical modeling applications (e.g., degradation of bio-resorbable stents). By advancing computational modeling and data assimilation, this research aims to bridge the gap between theoretical simulations and clinical practice, facilitating the development of more effective, personalized, and safer ablation therapies.

Background: Pulsed field ablation (PFA) is a new non-thermal ablation method being rapidly adopted for treatment of cardiac arrhythmia. The goal of PFA is to achieve sufficient electric field strength in the target region to produce the desired lesion depth while minimizing thermal damage.

Objective: Use numerical modelling to determine the lethal electric field threshold (LET) and thermal damage for PFA in healthy ventricles and to predict the extent of PFA lesions in infarcted ventricles.

Methods: We used fifteen 40-50 kg Yorkshire swine (10 healthy, 5 infarcted) in the study. PFA was performed using a two-catheter setup: a focal ablation catheter in the left ventricle (LV) and a return catheter in the inferior vena cava (IVC). In healthy ventricles, the pulse amplitude (1000-1500 V) and the number of pulse trains (1, 4, 8, and 16) were varied. In infarcted ventricles, a fixed protocol (1500 V, 8 trains) was used.

The numerical models were developed in COMSOL Multiphysics (v6.1).

To model PFA in healthy ventricles we used schematic biventricular geometry, constructed based on MRI, accounting for cardiac tissue anisotropy. LET was determined by calculating the electric field distribution for each lesion location and determining the threshold at which the predicted lesion volume matched the volume observed on late gadolinium enhancement (LGE) MRI. To calculate temperature increase we first calculated the blood flow in the LV and IVC for a high and low blood flow condition in a stationary study and use the solution in the coupled electro-thermal time-dependent study. Potential thermal damage was assessed using the Arrhenius integral and a thermal dose threshold of 1 s at ≥ 55 °C.

To model PFA in infarcted ventricles pre-ablation LGE MRI were processed using ADAS 3D software to obtain animal-specific ventricular geometry and scar tissue distribution, which was used to assign different electrical conductivities to the healthy myocardium and scar tissue in the model. Stationary study was used to determine the effect of scar tissue on the electric field distribution in LV.

Results: We determined LET for each PFA lesion in healthy tissue at 24 hours, 7 days, and 6 weeks post-ablation. Median LET varied with the number of pulse trains; at 7 days, thresholds were 725, 520, 484, and 394 V/cm for 1, 4, 8, and 16 pulse trains, respectively. Median LET was lower at 24 hours than at 7 days and 6 weeks, reflecting larger lesion volumes observed on LGE MRI at the earlier time point.

The maximum predicted thermal damage volume was minimal - 13 mm³ (less than 2% of total lesion volume) for the highest dose pulse protocol (16 trains) under low blood-flow condition - consistent with the absence of a thermal signature on MRI.

In infarcted ventricles, numerical simulations demonstrated PFA’s ability to create lesions through dense scar tissue. Predicted lesion extents agreed closely with histology.

Conclusion: We determined the LET for ventricular tissue in vivo, showing its dependence on the number of pulse trains and the time of evaluation. Numerical modelling and MRI confirmed the non-thermal nature of our PFA protocols.

In the landscape of thermal ablation strategies, laser ablation (LA) has emerged as a minimally invasive approach that employs focused laser energy to induce hyperthermia, effectively destroying pathological tissue. This mechanism, based on a rapid temperature increase, has demonstrated high spatial selectivity. While LA has been widely used in oncology, its potential for treating cardiac arrhythmias by targeting ectopic foci is gaining increasing attention. Laser light is absorbed by the tissue and converted into heat, resulting in localized thermal damage to the targeted area. Despite its clinical promise, predictive modeling tools for LA in cardiac applications remain limited, especially when compared to the well-developed models available for radiofrequency ablation.

Temperature distribution plays a pivotal role in determining lesion size and therapeutic efficacy. Small deviations in thermal delivery may lead to incomplete ablation or collateral damage to nearby healthy tissue. In cardiac tissue, the anisotropic and heterogeneous structure further complicates the modeling of heat propagation. Moreover, the optical and thermal properties of myocardial tissue significantly affect energy deposition patterns, making accurate, physics-informed computational modeling essential for virtual planning and outcome prediction.

This study proposes a comprehensive numerical framework tailored for simulating laser-tissue interactions in cardiac environments, aiming to enhance in silico planning capabilities for interventional procedures. A three-dimensional idealized cardiac tissue domain is modeled, incorporating rotational anisotropy and simulating the fully coupled optical–thermal response using a custom finite element implementation. The framework evaluates and contrasts several heat conduction models, starting with the classical Pennes bioheat equation and extending to more advanced formulations, including the Generalized Fourier (GF) model and the Dual-Phase Lag (DPL) model. These higher-order approaches account for finite thermal propagation speed and capture microstructural effects often neglected in traditional Fourier-based models.

To assess cellular damage, the thermal models are integrated with a multiscale three-state cell death dynamics model, which distinguishes between healthy, reversibly injured, and irreversibly damaged cells. This allows a more nuanced evaluation of ablation outcomes compared to conventional thresholds such as the 50 °C isotherm. Parametric analyses reveal that higher-order thermal models provide a more realistic prediction of lesion development, especially under fast heating conditions.

The results underscore the critical role of cardiac tissue anisotropy and validate the potential of advanced thermal modeling in improving lesion prediction accuracy. These findings contribute to the advancement of digital twin technologies in cardiology, offering a foundation for future integration of LA protocols into personalized simulation platforms aimed at guiding clinical decision-making in the treatment of cardiac arrhythmias.

Patients with ischemic cardiomyopathy and previous myocardial infarction often develop regions of fibrotic scar tissue in the heart, which can create a structural environment that supports reentrant ventricular tachycardia (VT) [1]. Accurately identifying these arrhythmogenic regions is essential to guide catheter ablation therapy effectively. Late gadolinium enhancement cardiac magnetic resonance imaging (LGE-CMR) allows for non-invasive visualization of myocardial scarring [2], while electroanatomical mapping (EAM) identifies low-voltage areas that indicate damaged or fibrotic tissue [3]. However, differences can exist between what is seen on MRI and what is detected by voltage mapping, due in part to anatomical factors like wall thickness. Recent work has shown that computational models based on MRI data can help identify patient-specific VT circuits and potential targets for ablation [4].

We developed a novel workflow to better characterize the arrhythmogenic substrate in patients with left ventricular myocardial infarction. LGE-CMR images were used to segment and classify scar regions into dense core and surrounding border zones, assigning different electrical properties—either non-conductive (dense scar) or with reduced conductivity (border zone)—based on MRI signal intensity. These data were then combined with 3D electroanatomical mapping of the left ventricle, obtained with the Carto 3 system, to merge structural and electrical information into a unified model.

Using the reconstructed geometry, we simulated electrical activation across the heart tissue through a simplified electrophysiological model [5], applying programmed stimulation at various points on the endocardial surface to test for reentrant circuit dynamics.

This study introduces a robust and integrated modeling approach that combines imaging, electro-anatomical mapping, and simulation to improve understanding of post-infarction VT. By linking structural and electrical information to a single model, this workflow offers a powerful tool for non-invasive identification of critical conduction pathways and potential VT circuits, with significant implications for improving the planning and precision of ablation procedures.

References

[1] R. Lo, K. K. M. Chia, and H. H. Hsia, «Ventricular Tachycardia in Ischemic Heart Disease», Card Electrophysiol Clin, vol. 9, fasc. 1, pp. 25–46, Mar. 2017.

[2] S. Kuruvilla, N. Adenaw, A. B. Katwal, M. J. Lipinski, C. M. Kramer, and M. Salerno, «Late Gadolinium Enhancement on Cardiac Magnetic Resonance Predicts Adverse Cardiovascular Outcomes in Nonischemic Cardiomyopathy: A Systematic Review and Meta-Analysis», Circ: Cardiovascular Imaging, vol. 7, fasc. 2, pp. 250–258, Mar. 2014.

[3] P. Compagnucci et al., «Recent advances in three-dimensional electroanatomical mapping guidance for the ablation of complex atrial and ventricular arrhythmias», J Interv Card Electrophysiol, vol. 61, fasc. 1, pp. 37–43, Jun. 2021.

[4] H. Ashikaga et al., «Feasibility of image-based simulation to estimate ablation target in human ventricular arrhythmia», Heart Rhythm, vol. 10, fasc. 8, pp. 1109–1116, Aug. 2013.

[5] A. J. Pullan, L. K. Cheng, M. L. Buist. Mathematically Modelling The Electrical Activity Of The Heart: From Cell To Body Surface And Back Again. 2005.

This study combines advanced numerical methodologies with additive manufacturing (AM) to optimize the design and production of orthopedic insoles. The research aims at reducing material usage and production time, while maintaining the mechanical properties required for optimal performance. This is accomplished by merging experimental evidence into computational modelling.

In recent years Additive Manufacturing (AD) has been successfully integrated in the design and production of orthoses, to the benefit of a customizable correction of impaired foot biomechanics. As well known, 3D printed patterns (or “infills”) consist in an alternation of void and full material regions, intrinsically allowing to save on material consumption. In particular, this work analyzes the widely-used honeycomb infill pattern, which, while mechanically robust, is associated with longer printing times. Alternative patterns have been considered to identify a suitable replacement infill that is mechanically comparable but is printed faster.

As a second step, the mentioned material savings have been further optimized in the insole fabrication by distributing the replacement infill only where it is actually needed. To accomplish this, the research as leveraged the potentialities offered by topology optimization, a numerical technique that determines the optimal material distribution within a structure to minimize weight while ensuring structural integrity. Specifically, the optimization process is applied to the frontal region of the insole, where material reduction is achieved while considering mechanical stability under real-world loading conditions. These include bending and deformation caused by the insertion into the shoe.

By addressing both structural and printing optimization, the final insole registers significantly reduced production time and material usage. The proposed design is a trade-off between portability, comfort, and usability, and demonstrates the advantages of integrating advanced computational techniques with additive manufacturing for biomedical applications.

This work is part of the Spoke 2, FP4 (partner UCBM) Rome Technopole project, funded by the E.U. and the Italian Ministry of University and Research.

In recent years, wearable devices for monitoring physiological and biomechanical parameters have attracted growing interest, particularly in the assessment of individuals suffering from low back pain. This condition can influence parameters such as respiratory rate (RR) and heart rate (HR), making their continuous and non-invasive monitoring highly valuable for the implementation of pain mitigation strategies. In this context, Fiber Bragg Grating (FBG) sensors have emerged as particularly promising technology due to their small size, multiplexing capability, and immunity to electromagnetic interference. An innovative solution to overcome the intrinsic fragility of FBGs, while enhancing their strain response, consists of integrating the sensors into components manufactured by 3D printing, an approach that offers cost-effectiveness, reproducibility, and ease of fabrication.

This study proposes the design and fabrication of two wearable sensors (WSs), identical in geometry but differentiated by function, based on FBG sensors encapsulated into components fabricated using dual-extrusion 3D printing with Fused Deposition Modelling (FDM) technology. One WS is designed to be positioned horizontally on the subject’s thoracic cage to monitor cardio-respiratory activity, while the other is intended to be placed vertically in the lumbar region to detect flexion-extension movements of the back. In particular, the cyclic movements induced by respiration and cardiac activity at the thoracic level, as well as lumbar flexion-extension movements, generate periodic strain variations on the WSs, resulting in corresponding shifts in the reflected wavelength of the integrated FBGs. The analysis of such variations enables the extraction of physiological and biomechanical parameters such as RR, HR, and flexion-extension rate. The main innovation introduced in the two WSs lies in the combined use of two materials: thermoplastic polyurethane (TPU), used to fabricate a dog-bone-shaped element integrating a single 10 mm-long FBG at its central portion, and polylactic acid (PLA), employed to fabricate two lateral flanges designed to provide structural rigidity and allow the anchorage of the device to the subject’s body.

Following fabrication, a metrological characterization was carried out on one of the two WSs to evaluate its temperature sensitivity (S_T) and strain sensitivity (S_ε), as well as its dynamic behavior under cyclic loading conditions. The results showed a value of S_T equal to 0.034 nm/°C and a S_ε of 0.81 nm/mε, in line with values reported in the literature for similar sensors. Moreover, the device demonstrated good repeatability and adequate responsiveness to dynamic variations.

Finally, preliminary tests were conducted on eight healthy volunteers to assess the effectiveness of the two WSs in monitoring cardio-respiratory activity and lumbar flexion-extension movements. The acquired data were analyzed using custom algorithms developed in MATLAB. The results showed high accuracy in the estimation of RR and HR, with mean absolute errors lower than 0.5 breaths/min and 2.8 beats/min, and mean absolute percentage errors below 3% for both parameters. Furthermore, the analysis of movement sequences confirmed the system’s ability to effectively discriminate between different execution rates of flexion-extension movements.

Future developments will focus on device miniaturization, the integration of multi-sensor functionalities, and validation on clinical populations suffering from chronic low back pain.

In light of the escalating occurrence and severity of wildfires, this study explores the feasibility of deploying wearable devices to monitor cardiac, respiratory, physical, and environmental parameters during operational wildfire suppression tasks. Wildland firefighting strategies typically include direct attacks—where operators extinguish flames at close range—and indirect tactics such as debris removal and hose support, performed at varying distances from the fire front. While most research in this field is confined to laboratory simulations, the physiological burden experienced by firefighters (FFs) during live fire scenarios remains largely unexplored due to technical and logistical constraints. This study addresses this gap by conducting a comprehensive field-based analysis with two teams of six FFs from the Italian National Fire Corps, involved in a structured and controlled fire simulation protocol.

The protocol, designed to emulate real firefighting conditions, included six consecutive phases: baseline rest (Stop 1), two bouts of running (Run 1 and Run 2), intermediate rest (Stop 2), active fire suppression (Fire), and final recovery (Stop 3). FFs were equipped with commercial wearable chest straps capable of capturing ECG, respiratory signals, and accelerometry, as well as external temperature loggers. These devices allowed for high-frequency recording of physiological metrics including heart rate (HR), heart rate variability (HRV), respiratory frequency (fR), and vector magnitude units (VMU) of physical activity. Data were processed with filtering and artifact-removal algorithms applied to ensure accuracy.

Results revealed intense cardiovascular and respiratory strain during active phases. HR often exceeded 85% of the age-predicted maximum in many subjects, surpassing this threshold by over 40%, reflecting considerable physiological stress. HR recovery (HRR) after intense bouts was heterogeneous across subjects; while some FFs showed rapid decreases in HR (HRR >12 beats per minute (bpm) within 60 s, indicating good recovery), others—especially those in debris management roles—showed sustained elevations, hinting at incomplete recovery. HRV decreased significantly in post-exertion phases, suggesting elevated sympathetic activity and limited parasympathetic reactivation.

Respiratory analysis, performed using a power spectral density (PSD) approach to counter motion artifacts, highlighted consistent episodes of mild tachypnea (fR > 20 respirations per minute (rpm)) after active fire suppression, particularly in high-temperature and high-exertion roles. A general return to eupnea (12–20 rpm) was observed during final recovery, indicating prolonged respiratory activation.

Moreover, the physical activity load, quantified via VMU, confirmed role-dependent intensity. Debris management roles showed continuous high-intensity physical engagement (VMU > 0.8 g for more than 40% of the fire phase), while hose handlers and support personnel remained predominantly in moderate to low ranges.

These preliminary findings confirm the reliability of integrating wearable technologies to monitor the physiological and physical strain associated with wildfire suppression. Despite minor signal disruptions during high-exertion tasks, the system proved robust under harsh conditions, capturing critical physiological and physical metrics. These results highlight the importance of role-specific monitoring, as task intensity and physiological demands vary considerably. By identifying key stress indicators, this approach provides a solid foundation for optimizing training protocols, refining work-rest strategies, and developing tailored interventions to enhance FFs’ safety and operational performance.

This paper is devoted to the multiscale modelling of transport processes in perfused tissues, such as liver, which are considered as multiporous media. The topic is important in the context of the perfusion diagnostics (in vivo), but also in the context of tissue engineering and regenerative medicine. In both these situations, the transport due to the advection-diffusion (A-D) is described at the tissue heterogeneity level, which can be treated in a dual way: 1) due to the hierarchical homogenization "micro-meso-macro", capturing the fluid-structure interaction (and accordingly also the A-D transport), or 2) using the "double porosity" ansatz which is based on a kind of "Double-Darcy model" (homogenization of the Darcy flow with large contrasts in the permeability and, accordingly, in the diffusivity and the advection velocity involved in the A-D transport equations. We consider both and compare both theses approaches in two situations:

Model-P: The contrast fluid (CF) enhanced CT perfusion test -- the A-D transport of the CF under the flow due to the standard blood perfusion;
Model-A: The CF (or other species) transport influenced by the acoustic wave propagation (with zero background flow) due the acoustic streaming (AS), which appears as a secondary effect of the acoustic waves propagating in the porous medium.

For Model-A, we consider periodic structures comprising solid skeleton and fluid-filled pores in which the CF is transported. Using the perturbation analysis, the nonlinear acoustics model is decomposed into linear, but coupled subproblems: the Acoustic Wave Propagation (AWP) problem, and the Acoustic Streaming Flow (ASF) problem. The two subproblems are treated by the asymptotic homogenization which yields the two-scale models for both these subproblems. The advection velocity field, as computed form the ASF subproblem, is involved in the A-D equation describing the transport of a dissolved species represented by the concentration. In the solid (or dual porosity - extra-vascular space, tissue parenchyma), a weak diffusion is considered, which The homogenization with the scaling ansatz is applied also to obtain the two-scale A-D transport model. A special care is devoted to the interface between the fluid channels (primary porosity, in general) and the dual porosity. We consider interface permeability and diffusivity (allowing for a jump in the pressure and concentration, respectively) which can depend on the ASF-generated shear stress). This enables to account for the "sonoporation" effect - the vessel wall becomes more permeable as the effect of the acoustic waves.

All the local and macroscopic models are implemented in the SfePy finite element code. Numerical examples illustrate the sensitivity of the delivery and transport on the microstructure geometry and other features.

Rohan, E., Naili, S., Homogenization of the fluid-structure interaction in acoustics of porous
media perfused by viscous fluid, Z. Angew. Math. Phys., vol 71, 137, 2020.

Rohan, E., Moravcová, F. Acoustic streaming in porous media – homogenization based two-scale modelling, 2024 J. Phys.: Conf. Ser. 2647 232009.

Rohan, E., Turjanicová, J., Lukeš, V. Multiscale modelling and simulations of tissue perfusion using the Biot-Darcy-Brinkman model, Computers & Structures, Volume 251, 2021, 106404.

This work presents a mathematical framework for modeling the transport of biological fluids and nutrients within vascularized tumours, incorporating microscale effects into a macroscopic representation. Tumour growth and progression are highly dependent on the exchange of fluids and nutrients with surrounding vasculature. However, the abnormal and heterogeneous nature of tumour-induced angiogenesis creates complex spatial variations in fluid flow and nutrient distribution, influencing the effectiveness of treatments such as chemotherapy and immunotherapy. Understanding these transport mechanisms is therefore crucial for improving predictive tumour models and developing more effective therapeutic strategies.

To capture these intricate dynamics, we employ asymptotic homogenisation, a widely-used upscaling technique designed to analyse and model multiscale systems. By exploiting the separation of scales, asymptotic homogenisation enables a systematic transition from detailed microscopic descriptions to simplified macroscopic models, while preserving essential microscale information encapsulated in effective parameters at a reduced computational cost. The model describes the tumour microenvironment as a double porous medium, consisting of both the tumour tissue and the embedded vasculature, which interact through fluid and nutrient exchange. The interstitial fluid, representing both extracellular fluids and cells, is assumed to be incompressible, with tumour growth accounted for through a volumetric source term proportional to nutrient availability and a non-convective mass flux driven by nutrient gradients. The governing equations describe fluid motion influenced by both pressure and nutrient concentration gradients, which reduce to Darcy’s law in absence of microscale variations. Nutrient transport is modeled through a coupled advection-diffusion-reaction system, where permeability and diffusivity tensors are derived from cell-level problems to account for microvascular geometry. The exchange of both fluids and nutrients between the tumour and vasculature is described using the Kedem-Katchalsky formulation, incorporating microscale transport properties into the macroscopic equations. This framework enables a computationally efficient representation of tumour-vasculature interactions while preserving critical microscale influences.

The proposed model serves as a powerful tool for simulating tumour growth and treatment response in a patient-specific context, as it can be adapted to realistic tumour geometries reconstructed from medical imaging. By integrating microscale vascular properties into macroscale tumour models, it enhances predictive accuracy and supports the development of personalized treatment strategies. While the current model assumes a static tumour domain and simplified proliferation mechanisms, it establishes a foundation for future advancements incorporating mechanical deformations, dynamic vascular remodeling, and therapy-induced modifications in the tumour microenvironment. These extensions will further bridge the gap between mathematical modelling and clinical applications, improving the model’s relevance for oncological research and treatment planning.

Modelling material failure is an open problem in continuum mechanics, with many modelling techniques filling the panorama of damage mechanics. Our research has focused on creating a model that can account for how microscopic changes effect the macroscopic process of material failure. To that end, we have developed a novel multiscale model of the damage phase field method. This was achieved via new analytical methods introduced for the upscaling of the damage phase field. As a result, we have a rigorous multiscale model of material damage, which can be employed to study a myriad of physical phenomenon. Including processes such as plasticity, cyclic damaging or sudden material failure. This work of course has many applications in the realm of material science, but we are more interested in clinical applications.

Our results were achieved three-fold. Firstly, we used the damage phase field method to develop a model of material failure across an arbitrary linear elastic composite material. Such a model allows us to approximate a wide variety of materials: including alloys, soft tissues and other engineered structures. Next, we employed asymptotic homogenisation to upscale our model into a computationally feasible macroscopic model. This macroscopic model encodes all the microscopic information about a materials microstructure into a set of effective coefficients. These effective coefficients are dependent on the solutions to a series of local cell problems which relate the macro and micro scales of our problem. Finally, we explored the question of how to solve our damage problem numerically. We found that a novel approach was required to calculate a solution. The damage phase field method has solutions depending on an optimisation problem which is not solved trivially. By developing a new numerical algorithm, we were able to implement effective code that allowed us to do two dimensional simulations of a trouser test. A trouser test is when we clamp a material in a rectangular reference configuration at one end and pull it apart at the other end. The resulting final configuration looks like a pair of trousers, hence the name. By studying the results of these simulations, we have found that a materials microscopic properties and geometry have a strong influence on material failure.

Our long-term goal is to apply our modelling methods to biological phenomena of soft tissue tearing. Namely, diseases such as aortic dissections and ACL ruptures. By applying a multiscale model of these diseases, we could potentially understand what microscopic changes in the body are making people more susceptible to these diseases. The clinical applications of these models would include predicting the occurrence of soft tissue tearing and preventing it, or a better understanding of the long-term effects of treatment. This would allow clinicians to make more informed decisions, leading to better patient outcomes.

Cells respond to their local environment through mechanotransduction, converting mechanical signals into a biological response, facilitating changes in cell function (e.g. cell growth, proliferation or differentiation). The cell cytoskeleton, particularly actomyosin stress fibres (SFs), and focal adhesions (FAs), which bind the cytoskeleton to the extra-cellular matrix (ECM), are central to this process, activating intracellular signalling cascades in response to deformation.
We present a novel bio-chemo-mechanical one-dimensional model to describe the formation and maturation of these mechanosensing structures, coupled through a positive feedback loop, and the associated cell deformation. In particular, we employ reaction-diffusion-advection equations to describe: the polymerisation of actin and bundling and activation of the resultant fibres; the formation and maturation of adhesions between the cell and substrate; and the activation of certain signalling proteins in response to FA and SF formation. A set of constitutive relations then connect the concentration of these key proteins to the mechanical properties of the cell cytoplasm and the ECM. In particular, we treat the cell as a Kelvin-Voigt viscoelastic material, with additional active stresses due to myosin II motor contractility. By connecting the cytoskeletal mesh scale to the microscale using discrete-to-continuum upscaling, our approach advances on the homogenised account of the cell provided by many existing models by rationally connecting the nanoscale and microscale features of cell-substrate adhesion and the cell cytoskeleton.
We employ this model to understand how cells respond to external and intracellular cues in vitro. For example, we explain how: dependent upon the mechanical properties of the surrounding ECM, non-uniform patterns of cell striation develop, leading to FA and SF localisation at the cell periphery; myosin II inhibition leads to disruption of SFs and, in turn, FAs; and nanoscale ligand patterning exerts control over microscale adhesion and cytoskeleton development. Having demonstrated the ability of the model to replicate experimental observations, the model provides a platform for systematic investigation into how cell biochemistry and mechanics influence the growth and development of the cell and associated changes in cell function, and facilitates prediction of internal cell measurements that are difficult to ascertain experimentally (e.g. stress distribution).

Finally, extending this model to two dimensions facilitates the incorporation of other key mechanosensing structures, including the stiff cell nucleus, and plasma and cortical membranes. With these additions, we consider the axisymmetric analogue of our one-dimensional model and conduct a linear stability analysis to investigate the stability of this axisymmetric configuration to various normal modes of deformation. By identifying non-axisymmetric modes with positive growth rates our model also reveals a possible mechanism for self-driven surface patterning of cells in vitro.

Lymph nodes (LNs) are essential components of the immune system, where lymph fluid, containing immune cells and antigens, is processed. Their structure consists of a porous lymphoid compartment (LC) and a surrounding thin subcapsular sinus (SCS) that allows free fluid flow. In this talk, we present a mathematical model, derived using the asymptotic homogenization technique, to capture the multiscale nature of fluid flow within the lymph node. We employ numerical simulations to investigate flow patterns, pressure distributions, and shear stress in detail. These results provide valuable insights into the mechanical environment of the lymph node, advancing our understanding of its role in immune function and offering a foundation for exploring therapies for lymphatic disorders.

Although the biochemical pathways of blood coagulation are well understood, the role of platelet mechanobiology in clot remodeling and the initiation of tissue repair remains less clear. Platelets not only secrete biochemical factors essential for the rapid formation of fibrin-rich clots, but they also contribute to wound contraction. Studies have demonstrated that platelets are instrumental in assembling the first fibronectin fibers, setting the stage for the subsequent migration of other cell types.

It is now widely accepted that cell motility is driven by the polymerization of actin—the most abundant protein in eukaryotic cells—into a network of interconnected filaments. We describe this mechanism within a continuum mechanics framework, suggesting that actin polymerization induces mechanical swelling in a localized region near the nucleation sites, which ultimately drives movement in cells or bacteria.

To explore the mechanobiology behind these phenomena, cutting-edge microscopy techniques [1,2] have been integrated with a novel theoretical approach, resulting in more sophisticated models and simulations. Departing from the commonly assumed incompressibility of all components in many existing mixture theories [3], we have extended the classical Larché-Cahn framework for chemo-transport-mechanics [4].

This new theoretical model has shown promising results in simulating cell motility [5] and could potentially be applied to other areas of mechanobiology as well.

References

[1] M. Burkhardt, et al. (2016). Synergistic interactions of blood-​borne immune cells, fibroblasts and extracellular matrix drive repair in an in vitro peri-​implant wound healing model. Sci Rep 6, 21071.

[2] S. Lickertet. al. (2022). Platelets drive fibronectin fibrillogenesis using integrin αIIbβ3. Science Advances, 8(10), eabj8331

[3] F.J. Vernerey and M. Farsad. A constrained mixture approach to mechano-sensing and force generation in contractile cells. J MECH BEHAV BIOMED, 4(8):1683–1699, 2011.

[4] M Arricca, L Cabras, M Serpelloni, C Bonanno, R M. McMeeking, and A Salvadori. A coupled model of transport-reaction-mechanics with trapping, Part II: Large strain analysis. J MECH PHYS SOLIDS, 181:105425, 2023

[5] A. Salvadori, C. Bonanno, M. Serpelloni, R.M. McMeeking, (2024), On the generation of force required for actin-based motility., Scientific Reports, 14:18384, https://doi.org/10.1038/s41598-024-69422-3.

Abstract: Cells can sense the mechanical properties, such as stiffness of extracellular matrix (ECM), and this mechanosensing capability can affect cell behaviors, tissue development, and pathological changes. Most ECMs exhibit dissipative mechanical properties, i.e., viscosity and plasticity. The dissipative properties of ECMs can regulate stem cell differentiation, cell spreading, and migration. However, the mechanism of the cell mechanosensing, especially sensing the dissipative properties, remains unclear. This report mainly introduces our exploration of the cell mechanosensing mechanism in recent years. Combining chemo-mechanical model development and biochemical experiments, we found that the contractile force generated by intracellular myosin can be transmitted to ECMs through binding dynamics of adhesion molecules. The competition between stress relaxation of the viscoelastic matrix and the adhesion binding dynamics determines the effective stiffness sensed by cells, thereby influencing the cell spreading and migration. Next, we studied how the ECM plasticity affects the dynamic behaviors of cancer cell invadopodia in 3D cell culture. We found that the competition between myosin contractility and actin polymerization drives the periodic oscillatory growth of invadopodia; during the invadopodia oscillation, the matrix plastic deformation gradually accumulates, forming a cavity in the invadopodial front and further promoting the invadopodia growth. In addition to the cancer cell invadopodia, we explored the mechanism of wave-like dynamics in immune cells’ podosome clusters, and our study shows that cells can utilize the podosome wave-like dynamics to sense the ECM stiffness. Overall, these studies reveal the previously unrecognized impact of the ECMs’ dissipative properties on cell dynamics, offering new clues for the design and optimization of biological materials.

Interaction of cells and their corresponding extracellular matrix (ECM) are known to regulate and control many different biological processes such as cancer progression and embryonic formation. A key component instructing cell-matrix interactions are the ECM’s material properties, i.e. its average, bulk response to cell generated stresses and strains. While cell-matrix forces as effectors of molecular signaling pathways and mechanical drivers of biological processes has been extensively studied for elastic ECM’s, native ECM properties are viscoelastic and characterized by including both elastic and dissipative components. Importantly, recent studies have shown that viscoelastic properties also moderate cellular behavior. Therefore, accounting for viscous properties in addition to the elastic properties is essential.

TFM is a widely established technique capable of quantifying cell-matrix forces by measuring strains applied on the ECM. Traditional elastic TFM algorithms (eTFM) make use of the analytical Boussinesq solution and Fourier analysis to provide a semi-analytical and computationally efficient algorithm for flat substrates with finite thickness. Most importantly, traditional TFM algorithms only assume that the substrate is elastic, neglecting the dissipative behavior of native and in vitro ECM. TFM approaches based on Finite Element models (FEM) can include viscoelasticity, but their numerical complexity and high computational cost restrict their practical application.

Here, we propose a viscoelastic TFM algorithm (veTFM) generalizing the semi-analytical elastic algorithm to viscoelastic substrates with finite thickness while maintaining the computational efficiency of traditional TFM algorithms. veTFM makes use of Fourier and Laplace analysis to reduce the complexity of the problem, while considering viscoelastic materials characterized by a two-component generalized Maxwell model. veTFM is validated in silico against analytical and FEM simulations, proving its validity in cases where the temporal variation of cell-generated strains on the ECM can be captured. Furthermore, veTFM’s applicability is experimentally evaluated in three different in vitro cases corresponding to muscle, epithelial and connective cell types. Comparing tractions obtained by veTFM to those obtained by traditional elastic algorithms shows that a viscoelastic ECM effectively behaves stiffer under rapid cellular force application and softer if forces are applied slowly. This confirms and quantifies previous observations on rate-dependent cellular response to viscoelastic substrates. Virtually extending this analysis to materials with a wide range of viscoelastic properties, including those of tissues, further shows that viscoelastic dissipation, i.e. the magnitude of stress dissipated by the viscoelastic ECM, in conjunction with the applied strain rate are the key factors differentiating viscoelastic from elastic traction force inference.

In conclusion, our method provides a practical, efficient framework for accurately characterizing how cells dynamically adapt their mechanosensing to viscoelastic mechanical environments, extending beyond the current methods restricted to elastic materials.

Notch is an evolutionarily conserved cell-cell signaling pathway central for several morphogenesis processes, such as cardiac development and angiogenesis. In humans, this pathway features several receptors (Notch1-4) and ligands (Dll1, Dll3, Dll4, Jag1, Jag2) that elicit different cell responses. For example, while the interaction of Dll4 with Notch1 inhibits the angiogenic response of endothelial cells, the expression of Jag1 promotes blood vessel formation instead. The presence of multiple ligands thus enables the tight control of morphogenic processes over time and space. When the ratio between the different ligands or their competition mechanisms are dysregulated, pathological morphogenesis arises. This is for example the case of cancer angiogenesis, characterized by an overexpression of Jag1 and exuberant blood vessel formation. Identifying the determinants of the effects of the different ligands and their competition can therefore increase our understanding of aberrant morphogenesis and lead to the development of appropriate therapeutic strategies. Recent studies indicate that a minimum amount of force needs to be exerted on the Notch receptor-ligand complex to elicit Notch activation. The role of this phenomenon in determining the ligands' effects is currently unclear.

Here, we propose a new computational model of Notch signaling to investigate the role of the force-mediated Notch activation for the different ligands' roles. An agent-based cellular automata approach was chosen, to account for the spatial features of the system in exam and to capture possible steric hindrance effects. The model featured a squared lattice representing the interface between two cells: one expressing Notch receptors (receiving cell) and the other expressing Notch ligands (signaling cell). These proteins could perform different actions at each time-step, mimicking the main characteristics of Notch signaling such as random protein movement across the cell membrane and protein endocytosis. When a Notch receptor and ligand co-localized on the same grid, the Notch receptor-ligand complex formed, followed by either Notch activation or protein dissociation. The likelihood of the two events in the stochastic model was assumed to be influenced by the catch-bond behavior characterizing the Notch protein complexes, as well as by the minimum force required for Notch activation.

A simplified version of the model with only one type of receptor and ligand was first validated against ordinary differential equations, usually adopted to model Notch signaling. The agent-based model was able to predict the same results of the ordinary differential equation. Next, the simulations were adopted to investigate whether differences in magnitudes of force necessary for Notch activation can explain differences in the roles of the Notch ligands. In agreement with literature, by assuming that a higher force is necessary to elicit Jag1-mediated Notch1 activation compared to Dll4, we were able to predict a stronger Notch1 response to Dll4 versus Jag1 exposure. These simulations therefore point at the magnitude of force necessary to induce the activation of a Notch complex as determinant for the role of Notch ligands. In future studies, the model will be extended to simulate cells expressing different ligands interacting among each other, thereby more accurately capturing their competition.

The migration of epithelial cells plays a critical role in physiological processes such as wound healing. In this context, cells utilize distinct migration modes based on the geometric properties of gaps: lamellipodial crawling at convex edges and purse-string-like movements at concave edges. Despite advances in identifying biochemical pathways, the underlying mechanisms determining these mode switches in response to curvature remain unclear. In several studies, we identified a critical link between the morphologies of the endoplasmic reticulum (ER), the mitochondria and the whole cell, and together, these morphologies can dictate various cell functions including cell migration and metabolic processes.

Specifically, through a combination of experimental data and theoretical modeling, we show that the ER undergoes curvature-specific morphological reorganizations that act as a determinant of migration modes. At convex edges, the ER forms tubular networks that align perpendicularly, facilitating lamellipodial crawling. At concave edges, the ER reorganizes into dense sheet-like structures favoring actomyosin-driven purse-string contractions. Our mathematical model describes the ER as a flexible fiber whose morphology-dependent strain energy guides these transitions, revealing a lower energy state when ER tubules or sheets form in accordance with local edge curvature.

This study positions the ER as a critical player in cellular mechanotransduction, providing a mechanistic link between subcellular organization and cellular migration strategies. Our findings offer insights into how cellular and subcellular geometries dynamically influence the physical properties and behaviors of cells, forming a basis for understanding migration regulation in complex tissues.

State estimation in large arteries presents a significant challenge, requiring robust methods that integrate physical models with experimental measurements such as Doppler ultrasound and 4D-flow MRI. In this study, we focus on reconstructing velocity, viscosity, and pressure states in shear-thinning blood flows using a projection-based approach. We compare three model order reduction strategies: classical Singular Value Decomposition (SVD), randomized SVD (rSVD), and a novel parametric Dynamic Mode Decomposition (DMD) method. The latter not only extrapolates partial velocity data beyond the acquisition region but also provides a predictive framework for super-sampling MRI data, which typically suffers from low temporal resolution.

Our methodology is validated in a patient-specific internal carotid artery siphon, where the state estimation is conducted within a manifold learning framework. Benchmark evaluations are performed for clinically relevant hemodynamic quantities, including pressure drops, oscillatory shear index (OSI), wall shear stress (WSS), and vorticity dynamics. Numerical simulations for the non-intrusive model order reduction are performed using the finite element method (FEM) for a non-Newtonian variation of the incompressible Navier-Stokes equations. A Carreau-like model is employed to account for varying hematocrit levels, while a variational multi-scale stabilization strategy ensures numerical stability in low-order elements.

Furthermore, we explore methodological advancements in inter-patient reconstructions. Specifically, we introduce an embedding strategy that enables state recovery in new arterial geometries without requiring a new offline training phase. This is achieved through a geometric mapping via a large diffeomorphism metric matching approach, enhanced by a multi-grid residual network for performance optimization. To ensure mass conservation, we employ a Piola transformation, while a nearest-neighbor search mechanism selects the most suitable pre-trained reduced-order model (ROM).

Our findings demonstrate the potential of combining model order reduction with data assimilation to improve the accuracy and efficiency of hemodynamic state estimation in clinical applications. This approach paves the way for real-time patient-specific simulations and enhanced diagnostic capabilities in vascular medicine.

Computational hemodynamics can enhance image-based diagnostic and provide complementary insights to predict, understand, and monitor treatments. The high computational costs and the complexity associated with handling patient-specific settings still remain a major challenge towards clinical applications. In this work, we propose a robust shape registration method for aortic geometries and its application to surrogate hemodynamic models and data assimilation tasks.

The approach is based on ResNet-LDDMM trained with a dataset of synthetic shapes, generated via SSM from an initial cohort of aortic coarctation patients and healthy subjects.

The loss function for training the ResNet is tailored to surface meshes, considering a modified Chamfer distance that accounts for mesh boundaries and orientations.

Moreover, the optimization utilizes a multigrid strategy, refining mesh size over the training epochs, that allows to handle realistic mesh sizes. The registration allows to define geometric encodings of different blood flow solutions on a single reference shape, as well as to build projection-based reduced-order models.

We employ this geometrical encoding to improve the training of encode-process-decode graph neural networks (GNN) and present potential applications in data assimilation problems, combined with a generalized Parametrized-Background Data-Weak formulation.

In partiular, we address the reconstruction of velocity fields and wall shear stresses, as well as the estimation of pressure fields and pressure-related biomarkers, like the pressure drop, from low-resolution velocity observations. We show different numerical tests based on the synthetic data, comparing the proposed strategies with state of the art estimators.

Accurate patient-specific blood flow simulations require precise specification of material parameters and boundary conditions to be a useful tool in clinical applications. Although no-slip conditions are commonly applied, allowing for some degree of slip at the vessel wall might, in many cases, lead to better agreement with flow patterns observed in medical images. In an attempt to address this issue, we are solving an inverse problem to estimate the boundary condition parameters, specifically focusing on the inlet velocity profile and wall slip parameters, using measured MRI velocity data (4D PC-MRI) obtained in descending aorta over one cardiac cycle.

The implementation is done using the Firedrake framework, an automated system for solving partial differential equations using finite element method, together with Pyadjoint, a library that enables efficient computation of adjoint equations by automatically differentiating the forward problem. Pyadjoint also contains an interface to various parallel implementations of optimization algorithms such as L-BFGS. These tools provide a flexible and efficient way to handle optimization problems in computational fluid dynamics, making them convenient for our application. Since solving inverse problems is usually computationally expensive, the forward problem is solved by the Newton or Picard method with a direct solver and discretized using P1P1 finite elements with interior penalty stabilization to reduce the overall cost.

To assess the robustness of the approach before applying it to real MRI data, we first conduct tests using artificially generated velocity fields under various conditions. This includes varying the amount of noise and resolution of the artificially generated velocity data, different numerical settings of the forward problem, and various regularization strategies of the error functional, which was done first for steady flow and then for transient flow.

The approach is also applied to multiple sets of real MRI patient-specific data, consisting of a morphological image (3D SSFP MRI) and velocity data (4D PC-MRI). Since patient-specific simulations usually contain some amount of uncertainty regarding the vessel wall, multiple segmentation of the morphological image are tested to study the sensitivity of the approach to this known phenomenon. Additionally, we compare the impact of Navier’s slip boundary condition (both constant and spatially varying) against the traditional no-slip condition, analyzing its influence on the flow simulations. By refining boundary condition estimations, this method can enhance post-processing analyses such as wall shear stress (WSS) calculations and pressure reconstruction, both of which are critical for assessing vascular health and disease progression. Improved flow accuracy may also lead to better estimates of hemodynamic biomarkers, further bridging the gap between computational modelling and clinical decision-making.

Internal carotid artery (ICA) stenoses are a major source of stroke. Besides typical cardiovascular risk factors, plaque composition, carotid bifurcation geometry, and hemodynamic parameters such as wall shear stress (WSS) are recognized as critical contributors to the progression and rupture of ICA plaques [1].

4D flow MRI provides time-resolved 3D velocity data for patient-specific hemodynamic assessment. Integrating this with CFD improves accuracy, especially in carotid stenosis, supporting risk stratification and treatment planning [2]. This study uses patient-specific CFD simulations to assess hemodynamics at the carotid bifurcation. The approach helps identify biomarkers for risk evaluation, personalized treatment, and stroke prevention across varying stenosis types.

Flow data and vascular geometries were obtained from 120 patients with mild carotid stenosis using 4D flow MRI from a previously published cohort [3]. These inputs enabled patient-specific simulations to comprehensively analyze hemodynamic variations and disease progression. Vessel segmentation was carried out using a nnU-Net model trained on semi-manual annotations by an experienced neurologist in MeVisLab.

Using VMTK, a centerline is generated for each geometry, with boundary patches constructed orthogonally along it. From phase-contrast images, 3D velocity vectors are extracted at these locations across all time points to define patient-specific inlet conditions. Appropriate meshes are using blockMesh and SnappyHexMesh utilities from OpenFOAM. CFD simulations are then run with validated solver parameters, followed by automated post-processing to derive relevant hemodynamic parameters.

This study focused on stenosed cases featuring plaques specifically located along the outer wall of the ICA. Key hemodynamic parameters included the pressure gradient, defined as the difference between maximum and minimum plaque surface pressure normalized by the projected centerline distance and the maximum WSS. Results demonstrated a consistent relationship between pressure gradient and maximum WSS, with elevated WSS often corresponding to increased pressure gradients. The disturbed flow was assessed using the oscillatory shear index (OSI) and relative residence time (RRT), where higher values indicated turbulence, recirculation, and fluctuations in WSS, conditions associated with plaque development and progression.

Stenosis severity alone does not determine hemodynamic risk; plaque morphology, location, and arterial geometry significantly influence localized forces. Certain geometrical features can lead to the co-localization of high-pressure gradients and WSS, increasing rupture risk. A steep reduction in cross-sectional area from the common carotid artery (CCA) to the peak stenosis in the ICA amplifies pressure gradients, particularly in steeper plaques, while flatter plaques exhibit lower gradients. Plaques extending into the CCA experience higher pressure gradients than those confined to the ICA. These findings highlight the critical role of arterial anatomy and plaque morphology in assessing plaque vulnerability.

A cerebral aneurysm is a bulge that develops from a weak spot in the wall of a blood vessel in the brain. The rupture of such an aneurysm often leads to severe disability or the death of thousands of patients every year. However, several endovascular treatment options exist to prevent vessel wall rupture. In these minimally invasive surgical procedures, a medical device such as a coil or a WEB (Woven EndoBridge) is inserted into the bulge, or blood flow is diverted using a stent-like device, called flow diverter. The goal of all interventions is to reduce blood flow inside the aneurysm to trigger coagulation, thereby achieving complete occlusion of the aneurysm sac.

To support clinical planning and treatment, for each individual patient, numerical simulation can provide detailed insights into the patient-specific setting and serve as additional information for the attending neuroradiologist. Therefore, accurate models of all the various devices [2], such as flow diverters or contour neurovascular systems within cerebral aneurysms, are essential. Especially in complex, patient-specific cases, the combined application of different endovascular devices may offer improved treatment outcomes and long-term stability by accommodating diverse anatomical conditions up front into the decision process. Each of these devices consists of multiple thin nitinol wires, where their final placement significantly impacts the success of the intervention. To assist in this implantation process, we outline how a virtual deployment can be realized with the open-source multiphysics framework 4C[1]. The wire-like device will be modelled with geomtrically exact Simo-Reissner beam finite elements[3]. Different options are available to model the behavior at intersections due to interacting beam geometries. The deployment process consists of multiple steps: First, the flow diverter is compressed into a microcatheter. Then, the microcatheter moves the device into the vicinity of its target location. Finally, the device is deployed and expands inside the vessel. Throughout these steps, a consistent mixed-dimensional coupling approach [4] is employed to account for interactions between the device and the microcatheter or arterial wall. Within each intervention step a finite element approach is used to account for the mechanical behavior of the device, wall and catheter ensuring accurate representation of the deformations and interaction until the device reaches the desired position. Within this presentation, we will detail our modeling and simulation workflow and will compare our method to existing approaches in literature. Furthermore, we will discuss how this virtual deployment may increase the success rate of interventions through patient-specific treatment strategies.
Keywords:
finite element method, mixed-dimensional structural modeling, virtual endovascular deployment

References:
[1] 4C. 4C: A Comprehensive Multiphysics Simulation Framework. https://www.4c-multiphysics.org. Accessed: 12.05.2025. 2025.
[2] Martin Frank et al. “Numerical simulation of endovascular treatment options for cerebral aneurysms”. In: GAMM-Mitteilungen 47.3 (2024), e202370007.
[3] Christoph Meier, Alexander Popp, and Wolfgang A. Wall. “Geometrically Exact Finite Element Formulations for Slender Beams: Kirchhoff–Love Theory Versus Simo–Reissner Theory”. In: Archives of Computational Methods in Engineering 26.1 (2017), pp. 163–243. doi: 10.1007/s11831-017-9232-5.
[4] Ivo Steinbrecher et al. A consistent mixed-dimensional coupling approach for 1D Cosserat beams and 2D surfaces in 3D space. 2022. doi: 10.48550/ARXIV.2210.16010. url: https://arxiv.org/abs/2210.16010.

Introduction Minimally invasive cardiovascular interventions have significantly improved patient outcomes across various pathologies. However, post-operative complications remain a concern in every procedure, often influenced by device selection, performance, and procedural conditions. The use of finite element (FE) simulations (in-silico) and machine learning (ML) models trained for the prediction of interventional procedures, can facilitate the process of prosthesis selection, pre-procedural planning and post-operational risk assessment. The main bottle neck of training a Machine learning model on clinical data is the data accessibility. In this study, we present a comprehensive framework for assessing Thoracic Endovascular Aortic Repair (TEVAR) procedures by integrating structural simulations, synthetic population generation, and machine learning (ML) models. We use the relevant clinical data and imaging to augment the population, using statistical shape modelling (Anatomic analysis), create individual digital twins (high fidelity virtual replicates) of the TEVAR procedure, assess the post-operational complications and train machine learning models to predict the outcome of this procedure. This approach aims to provide additional information for clinicians in prosthesis selection and predicting procedural outcomes, ultimately reducing post-operative risks and enhancing intervention success rates.

Methods The framework utilizes statistical shape modeling to generate a synthetic population of thoracic aortas based on computed tomography (CT) scans from 25 patients who have undergone TEVAR. This original patient population is augmented by 100 synthetic anatomies, using statistical shape modelling. In this process, the geometrical characteristics of the real patients are assessed. Using principal component analysis, the most prominent shape modes are selected. Then assigning different weights to these shape modes new geometries, with similar geometric characteristics are created. Then statistical analysis on anatomical and pathological data of TEVAR patients are conducted to evaluate the representativeness of the synthetic geometries. In the second step, a simplified simulation model is employed to reduce computational costs while maintaining predictive accuracy. The results are compared with previously validated TEVAR simulation models. In the last step, different ML models are trained to determine the landing zone of the device for surgery planning.

Results The synthetic cohort of 100 virtual cases is analyzed, demonstrating that the generated geometries preserve key morphological characteristics of the initial patient-specific anatomies. The population exhibits favorable compactness, specificity, and generalization, enabling a thorough evaluation of post-operative device performance. The simplified simulation approach reduces computational time by 60% while maintaining prosthesis configuration deviations under 10% compared to standard models. Furthermore, the ML-based predictive model achieves 80% accuracy in determining the device landing zone.

Conclusions The integration of in-silico modeling with machine learning-driven intervention simulations provides clinicians with a powerful tool to refine prosthesis selection, streamline pre-procedural planning, and enhance post-operative risk assessment. By improving precision and reliability in cardiovascular interventions, these advancements ultimately lead to better patient outcomes.

Identifying carotid atherosclerotic lesions at risk of plaque rupture, a key precursor to cerebral embolism and stroke, is of critical clinical importance. High stress within the fibrous plaque cap has been proposed as a risk factor; however, the influence of residual strains on stress distribution remains poorly understood.

We present an advanced computational framework integrating homeostasis-driven Growth and Remodeling (G&R) into patient-specific carotid geometries to predict residual strains and assess their impact on plaque stress. The method employs a multiplicative kinematics-based remodeling approach to homogenize tissue stress while accounting for heterogeneous plaque composition. Patient-specific vessel reconstructions were generated from Computed Tomography Angiography (CT-A) imaging, with tissue classifications obtained via an artificial intelligence (AI)-driven histology-based segmentation tool, distinguishing vascular matrix (MATX), calcifications (CALC), lipid-rich necrotic core (LRNC), and intra-plaque hemorrhages (IPH). The framework was applied to a cohort of 18 patient-specific cases, incorporating a total Lagrangian formulation and novel post-processing metrics for biomechanical evaluation.

The incorporation of residual strains reduced peak wall stress by up to 30%, though the extent of reduction varied depending on plaque morphology and tissue composition. While G&R consistently reduced stress peaks within matrix-rich regions, high calcification led to localized stress concentrations, restricting remodeling. Additionally, the effect of residual strains on stress distribution was found to be negligible in cases with either minimal or severe stenosis, underscoring the role of plaque heterogeneity in determining biomechanical risk.

Our findings highlight the importance of patient-specific biomechanical modeling in evaluating plaque rupture risk. By incorporating residual strains and remodeling mechanisms, this approach provides a more physiologically relevant assessment of fibrous plaque cap stress, offering enhanced predictive capabilities for stroke risk stratification and potential clinical decision-making applications in atherosclerotic disease management.

Articular cartilage serves an important mechanical function in the human body, i.e., load bearing and shock absorption. For the design of implants for cartilage repair after injury or disease, it is key to thoroughly understand the unique biomechanical properties of the native tissue and potential substitute materials.

Here, we characterize the macroscopic large-strain mechanical properties of healthy and osteoarthritic human articular cartilage and perform histological investigations to determine their relation to the underlying microstructure. Furthermore, we use multimodal mechanical testing methods combined with hyperelastic nonlinear continuum mechanics modeling, finite element simulations, and immunohistochemistry to correlate the macroscopic behavior to the underlying microstructural properties of bioprinted cartilaginous constructs generated from bone marrow-derived mesenchymal stromal cells (MSCs) embedded in thiolated hyaluronic acid (HA-SH) hydrogels.

We individually characterize human articular cartilage tissue from the medial and lateral sides of the femoral condyle and tibial plateau to determine region‐specific mechanical properties. Our results show that there are no significant differences between the medial and lateral sides, but tissue from the tibial plateau is slightly softer than tissue from the femoral condyle. Osteoarthritis leads to a significantly softened mechanical response, which is related to corresponding microstructural changes. Furthermore, we confirm that the cartilaginous tissues based on chondrogenically differentiated MSCs in HA-SH hydrogels exhibit similar mechanical characteristics including nonlinearity, hysteresis, conditioning, and stress relaxation as human articular cartilage tissue. Our study reveals a positive correlation between the amount of the main components of the extracellular matrix (ECM) of articular cartilage, collagen (COL) and glycosaminoglycans (GAG), and the classical shear modulus as a measure of stiffness.

The presented results will help guide the design of implants that are able to restore cartilage structure and function, bridging biomechanics and regenerative medicine for osteoarthritis treatment.

Diabetes is a chronic condition characterized by hyperglycemia resulting from insulin deficiency or some physiological dysfunction. Current therapeutic strategies predominantly rely on external insulin administration through active syringe injections (SIs). Frequently however, Continuous Subcutaneous Insulin Infusion (CSII) are resorted to if an automized administration is needed. Despite their theoretical efficiency in managing blood glucose levels, both these methods present counter effective drawbacks, including the development of lipodystrophies (LPDs). LPDs consist in a pathological reorganization of the adipose tissue in response to an external fluid in the subcutaneous tissue and to the injection pressure, either in the form of lipoatrophy (tissue loss) or lipohypertrophy (tissue accumulation).

Some currently commercial fixed-site devices, including Subcutaneous Insulin Injection Ports (SIIPs), mitigate certain drawbacks of SIs, such as needle phobia and distress, while facilitating clearer application sites. However, the limitations associated with LPDs remain, especially because there is an increased development risk with repeated injections at fixed locations. LPDs emerge therefore as a critical concern, leading to alterations in adipose tissue morphology, hindering drug diffusion and ultimately leading to metabolic imbalances and higher insulin usage.

This study addresses the need for advanced modeling frameworks to optimize insulin port designs and mitigate LPD-associated inefficiencies. A multi-field computational approach is developed, using fluid dynamics simulations for realistic insulin injection scenarios, pharmacokinetics coupling to analyze insulin diffusion pathways, and image-based histological reconstructions to incorporate tissue damage models. The spatiotemporal dynamics of insulin distribution are evaluated under three scenarios: a Nominal Case (NC), where the Worst Case (WC) involving obstructions, and Optimal Case (OC) with enhanced device design.

The presented results will show how our modeling framework is capable of assessing tissue damage and insulin absorption delays because of successive injections. Insights into this “wall effect” created by damaged adipose has highlighted the importance of optimizing the geometry of the device. By addressing several realistic complications, the optimized devices enable the redistribution of potential damage across a wider tissue volume and a minimization of adverse effects.

The proposed framework sets the stage for improved insulin port device engineering, advancing therapeutic efficacy and achieving more stable and effective treatment outcomes for diabetic patients.

Introduction - Percutaneous coronary intervention (PCI) is the recommended procedure for the treatment of flow-limiting stenoses in obstructive Coronary Artery Disease (CAD). In the case of complex lesions, such as with diffuse and/or multivessel disease, it is not trivial to determine which and how many lesions need to be treated to achieve optimal restoration in myocardial blood flow at the tissue level. In this work, we integrate the cardiac and coronary 3D anatomy with pre-revascularization perfusion imaging into our in-house multiscale perfusion model to build a digital twin, which is then used to determine the optimal revascularization strategy through the simulation of a virtual PCI. We include a validation on 6 patients who underwent PCI and follow-up perfusion imaging, which is used as a quantitative benchmark.

Methods - 3D geometries of coronaries and myocardium are segmented from pre-treatment CT images. Perfusion simulations are run with a multiscale finite element model featuring 3D fluid-dynamics equations for blood flow in the epicardial arteries, coupled with multi-compartment Darcy equations for flow in the microcirculation. Coupling conditions are imposed at the interfaces to enforce mass conservation and force balance, and the myocardium is subdivided so that each epicardial branch supplies blood to its corresponding perfusion territory. Parameters in the computational model related to microvasculature are calibrated to reproduce the pre-revascularization map of myocardial blood flow (MBF) clinically obtained through a dynamic stress CT Perfusion protocol (dyn-CTP). Virtual PCI is performed through the geometric alteration of the epicardial arteries at the sites of the treated lesions, restoring lumen patency with a good angiographic result. The method is validated on 6 patients by performing the virtual replica of the exact same revascularization clinically performed and comparing model predictions with follow-up dyn-CTP images. Validation metrics include: qualitative evaluation of the evolution of perfusion defects and quantitative assessment of ischemic mass, defined as myocardial mass showing MBF < 101 ml/min/100g.

Results - The proposed microvasculature calibration allows to successfully reproduce pre-treatment perfusion maps with high spatial resolution and <5% error in ischemic mass quantification. Virtual PCI results correctly reproduce the regression of ischemia associated to all the treated lesions. The predicted reduction in ischemic mass is consistent with what observed from follow-up perfusion images, with an average error of 2.1%. The possibility of insufficient restoration in MBF, due to an incomplete revascularization, is also captured by the model in the patients where this occurred.

Conclusions - The proposed digital twin shows high accuracy in predicting both complete and incomplete regression in ischemic mass, holding great potential for pre-treatment planning regarding the optimization of PCI procedures. The inclusion of other factors, such as drug effects, inflammation and disease progression or regression at the epicardial level could push its use further, improving diagnostic performances of non-invasive exams and leading to highly targeted treatment options, with overall better management of patients affected by CAD.

Introduction

Mitral regurgitation is one of the most prevalent valvular diseases worldwide, causing a backflow of blood during systolic closure. Percutaneous treatment, a new alternative to open-heart surgery, can lead to fewer post-intervention complications compared to traditional intervention. Transcatheter Edge-to-Edge Repair (TEER) has become the most commonly used percutaneous repair technique in recent years. Based on the Alfieri stitch, TEER method consists of suture-like clipping of the two leaflet segments using dedicated devices (Mitraclip, Abbott, USA and Pascal, Edward LifeSciences, USA). Due to the novelty of these devices, their long-term biomechanical consequences remain unknown. The aim of this study is to evaluate the hemodynamic and biomechanical effects of the TEER devices compared to the Alfieri stitch.

Methods

Micro-CT scan (NanoScan PET-CT, Mediso) was used to create a mitral valve (MV) model based on comercially available Lifelike MV (BioTissue Inc., Canada). The 3D printed MV mold was filled with two silicones that have previously been validated (EcoFlex00-50 and DragonSkin10, Smooth-On Inc., USA). Mitral chordae were created using multiple debraided polyester strings. TEER devices (Mitraclip NT and Pascal Ace Implant) and Alfieri stitch were placed in A2P2 configuration. A left-heart double activation simulator was used under the following conditions: Heart rate: 70 bpm, Stroke Volume: 70 ml, Mean Aortic Pressure: 100 mmHg and Fluid viscosity: 3.9 cP. Hemodynamic results were obtained using transthoracic echography (iE33, Philipps Healthcare, USA) and analyzed using vector flow mapping. Two high-speed video cameras and Digital image correlation software (VIC 3D, Correlated Solutions, USA) were used to evaluate strain.

Results

Under physiological conditions, Alfieri's stitch induced the lowest Mean Pressure Gradient (4.13 ± 0.19 vs. 6.87 ± 0.21 and 6.29 ± 0.22 mmHg) and the highest Effective Orifice (2.15 ± 0.09 vs. 1.63 ± 0.09 and 1.74 ± 0.08 cm²) when compared to Mitraclip and Pascal devices respectively (p<0.001). Concerning the biomechanical consequences, major principal strain (E1) was distributed in homogenous pattern for the Alfieri technique while the Mitraclip and Pascal devices induced higher strain localized in the coaptation line (0.05 ± 0.02 vs. 0.09 ± 0.03 and 0.06 ± 0.02 respectively). Moreover, higher values of E1 were induced by the transcatheter devices (p<0.001). When analyzing the ventricular flow streamline at both early and end-diastole, Alfieri’s stitch induced more vorticity than the TEER devices which might result in a more efficient blood filling and ejection.

Conclusion

It was found that the Alfieri stitch was more effective at preserving ventricular filling and ejection efficiency than TEER devices, as well as reducing systolic strain on the leaf surface. Although Mitraclip and Pascal had a 30% difference in E1, both TEER devices exhibited similar hemodynamic behavior.

Stroke is the second leading cause of mortality among cardiovascular diseases. Mechanical thrombectomy (MT) has revolutionized the treatment of acute ischemic stroke by allowing the removal of the clot from occluded intracranial arteries. However, the success of MT is influenced by multiple factors, including the mechanical properties of the clot, stent retriever (SR), catheter, and vessel wall. Understanding these interactions is crucial for optimizing device design and improving patient outcomes. In silico trials- computational simulations of medical interventions- offer an insightful approach to systematically investigate these factors under controlled working conditions and the geometric/mechanical properties of devices used.

Past studies have investigated the outcomes of thrombectomy using patient-specific anatomy but have assumed the vessel wall to be rigid (Luraghi et al., 2021b, 2021a). This study presents an in-silico framework to analyze the impact of flexible catheter and vessel wall properties on MT performance. A high-fidelity finite element model (FEM) of a thrombectomy procedure is developed, incorporating realistic cerebral vasculature model, human-analog clot properties derived from experimental stress-strain tests, and flexible catheter/arterial wall behavior. The computational model captures the dynamic interaction between the catheter, clot, and vessel wall to evaluate critical performance metrics such as clot retrieval success, vessel deformation, and catheter kinking behavior in acutely curved vessels.

A parametric study is conducted to assess how variations in catheter flexibility and vessel wall stiffness affect the overall efficacy of MT. Previous research used tracking of a catheter via a predefined path along the vessel, but this work used realistic pushing methods that mimic an in vivo scenario. Catheter flexibility influences navigation efficiency and clot engagement, while vessel wall properties determine the extent of vessel deformation and possible damage during retrieval. Simulation results indicate that optimal catheter flexibility enhances clot retrieval efficiency while minimizing excessive forces on the vessel wall.

The findings from this study provide valuable insights into the mechanical aspects of MT, contributing to the development of improved catheter designs and patient-specific treatment strategies. Using in silico trials, clinicians and medical device manufacturers can improve thrombectomy techniques and optimize catheter properties to enhance safety and efficacy. Future research will focus on integrating patient-specific blood flow dynamics and thrombectomy device variability to further improve the predictive capabilities of computational models. This study proposes the potential of in silico trials as a cost-effective and ethical alternative to extensive in vivo and in vitro testing. The proposed framework paves the way for precision medicine in neurointervention, where device selection and procedural strategies can be tailored based on individualized vessel and clot characteristics.

References

Luraghi, G., Bridio, S et al., 2021a. The first virtual patient-specific thrombectomy procedure. J Biomech 126.

Luraghi, G., Rodriguez Matas, J.F. et al., 2021b. Applicability assessment of a stent-retriever thrombectomy finite-element model: Stent-retriever thrombectomy FE model. Interface Focus 11.

Transcatheter Aortic Valve Implantation (TAVI) is a minimally invasive technique for aortic stenosis treatment. Originally introduced for elderly patients at high surgical risk, TAVI is currently becoming the first-choice therapy even in low surgical risk, younger patients [1]. In this context, there is a need to assess the long-term performance of the bioprosthetic valves used for TAVI. The main limiting factor to the durability of TAVI valves is Structural Valve Deterioration (SVD). SVD is an irreversible degenerative process that can ultimately lead to the calcification of the implanted valves, but its underlying mechanisms are still incompletely understood [2]. This computational retrospective study aims to investigate the relationship between early post-TAVI aortic blood-dynamics and SVD. We build on a previous study [3], by further personalizing our computational model, to propose computational risk scores that correlate with a premature onset of SVD.
We consider patients who received a 23mm Edwards SAPIEN valve. The study population is made of two groups: patients with and without SVD at 5-10 years follow-up exam. We reconstruct patient-specific aortic geometries from pre-operative clinical images and we create post-TAVI scenarios by virtually inserting a bioprosthetic valve model. The bioprosthetic leaflets in open and closed configuration are obtained with a mechanical simulation starting from an idealized geometry, which mimics the SAPIEN valves’ leaflets design. Using the Finite Element library lifex (https://lifex.gitlab.io/), Computational Fluid Dynamics (CFD) simulations are performed in such virtual scenarios. We impose an inlet flow rate condition personalized by reconstructing, from Pulsed Wave Doppler images, the patient-specific velocity temporal evolution in the left ventricular outflow tract and rescaling its magnitude to match the cardiac output measurement. The bioprosthetic leaflets are implicitly represented in the numerical simulation as an immersed surface and the opening/closure dynamics are imposed a priori. The numerical results are then post-processed to find hemodynamics indices able to discriminate between the SVD and non-SVD groups.
The virtual insertion of the bioprosthetic valve model inside the pre-operative reconstructed geometries was qualitatively validated exploiting available post-TAVI medical images showing a good agreement between the position and orientation of the implanted stent and those of the virtual stent. The CFD results showed that the presence of the TAVI valve highly influences aortic hemodynamics, characterized by a high velocity jet in the ascending aorta and vortical structures around this jet. Depending on the patient-specific geometry and blood flow features, the evolution of the jet and vortices generates shear stress patterns on the aortic wall and bioprosthetic leaflets that are used to formulate the computational risk scores.
The results of this study suggest that post-TAVI blood dynamics may have an influence on the development of SVD. Moreover, the proposed risk scores could potentially assist clinicians in a patient-specific planning of follow-up exams, moving toward a personalized care.
[1] Cesario et al, J Clin Med, 13(20):6123, 2024.
[2] Kostyunin et al, J Am Heart Assoc, 9(19):e018506, 2020.
[3] Crugnola et al, CMPB, 259:108517, 2025.

Radiofrequency ablation (RFA) is an effective treatment of atrial fibrillation (AF), with pulmonary vein isolation (PVI), electrical isolation of the pulmonary veins from the left atrium, being the typical ablation strategy for this arrhythmia type. During the procedure, an electrode at the catheter tip delivers RF current at 500 kHz to ablate arrhythmogenic tissues. The RF current flows between the electrode and a dispersive patch (DP) placed on the patient's skin, typically on the back or thigh.
While a generally safe treatment, clinical studies show that female patients undergoing ablation for AF face higher complication rates than males, despite similar mortality outcomes. These disparities may arise from anatomical differences, potentially affecting tissue response during ablation. Investigating these variations could enhance patient-specific treatment strategies and improve procedural safety.
To explore these differences, we developed 3D in-silico models derived from a male and a female patient imaging data. The model geometries incorporate detailed anatomical structures, each with distinct conductivity properties. This study aims to determine whether male and female anatomical differences influence PVI outcomes, by analyzing the power distribution, the driver for heat generation in the tissues.

It has been recognized in the field of Psychiatry that different disorders might share some features, in terms of symptomatology. The advent of affordable genetic sequencing, and the use of statistical learning on large datasets, revealed that there are strong correlations between certain disorders; and the idea of the general neuropsychiatric risk factor that may increase an individual's vulnerability to developing specific disorders was reinvented and garnered a lot of attention recently.

One of the most striking comorbidities confirmed to have underlying common features are depression and anxiety. Clinicians who tried to elucidate this ‘couple’ sometimes claim that they are difficult to differentiate. Indeed, early-onset anxiety in youth often correlates with an elevated risk of developing depression later in life, and the inverse holds true as well. This overlap presents a significant clinical challenge, as it complicates therapeutic strategies at a time when rates of depression and anxiety are consistently rising in last decade, particularly among younger populations.

From the perspective of Physiological Complexity (pathological decomplexification of electrophysiological signals in human physiology), there is yet another common feature characteristic for this ‘couple’: electroencephalograms (EEG) and electrocardiograms (ECG) recorded in persons diagnosed with depression and/or anxiety are exhibiting characteristic levels of complexity different than in healthy controls (HC). In existing literature on this topic many researchers repeatedly demonstrated that based on fractal and nonlinear measures signals recorded in patients are clearly separable from HC. However, these signal patterns reflect deeper dysregulations; aberrated dynamics of the autonomous nervous system (ANS). Such dysregulation manifests as impaired capacity for emotional and physiological self-regulation, such as the inability to disengage from negative stimuli. The disruption of regulatory feedback mechanisms—particularly the interplay between cardiac activity and respiration—produces measurable physiological outputs that can serve as biomarkers for detection, disease severity monitoring, and assessment of recovery potential. Among consequences of this characteristic aberration of ANS function, might be an increased cardiovascular risk (CVDr). That can be observed in particular in women who hold higher risk of depression, and whose CVDr are often underdiagnosed leading to bleak statistics of majority of women dying after the first adverse cardiac event.

The primary objective of this research is to integrate diverse candidate biomarkers—including ECG-derived complexity measures (via wearable devices), accelerometer-based activity metrics, and validated levels of non-ceruloplasmin copper in serum (also measurable using innovative portable technology)—into a unified fuzzy logic model. This model aims to characterize a general risk factor encompassing both heightened cardiovascular vulnerability and the severity of comorbid depressive and anxious states. Through mathematical modelling, we seek to isolate clinically relevant variables that accurately reflect the individual’s affective state and stage of illness, providing actionable insights for clinicians engaged in diagnosis and personalized treatment planning.

Wearable sensors for pulse wave monitoring have gained significant attention in recent years due to their potential for continuous, non-invasive cardiovascular assessment. Despite advancements in this field, existing technologies such as photoplethysmography (PPG) face notable limitations, including susceptibility to light intensity fluctuations, skin tone variability, and dependency on sensor positioning. This work presents the design, development, and metrological characterization of a novel magnetic-based soft sensor for pulse wave monitoring that addresses these challenges while enabling standardized measurement protocols.

The proposed sensing system consists of a Hall effect sensor and a deformable silicone matrix encapsulating a permanent magnet. Its operating principle relies on detecting magnetic field variations caused by arterial pulsations, which displace the magnet relative to the sensor. Unlike optical approaches, this mechanism directly measures mechanical deformations while simultaneously monitoring contact force during sensor placement, a critical parameter often overlooked in conventional systems.

Metrological characterization showed a mean sensitivity of 0.36 V×mm^-1 across the measurement range, increasing to 0.68 V×mm^-1 within the 4 mm - 5 mm compression displacement range, where the response is nearly linear. In this last range, dynamic testing under simulated heart rate conditions (60 bpm -120 bpm) validated the sensor's robustness during cyclic loading, with hysteresis errors ranging from 5.2% to 7.4% depending on loading rate. The minimum detection limit was established at 0.05 mm, indicating the sensor's ability to detect subtle arterial pulsations.

Feasibility assessment on eight healthy volunteers confirmed the system's capability to detect systolic, diastolic, and dicrotic peaks compared to a reference PPG sensor. Analysis of peak detection accuracy revealed promising performance, with systolic peaks showing only 0.10% false positives and 0.20% false negatives compared to the reference. The dicrotic peak demonstrated similarly strong results (0.08% false positives, 0.26% false negatives), while diastolic peak detection presented the greatest challenge with 2.79% false positives and 0.67% false negatives. Using the detected systolic peaks, beat-to-beat heart rate was calculated and compared with that derived from the reference PPG sensor, achieving a mean absolute error of 0.6 bpm across all subjects.

A key advantage of our sensor is its ability to measure contact force during positioning, enabling standardized signal acquisition and improved reproducibility. Tests evaluating the effect of sensor tightness confirmed that appropriate contact pressure is critical for high-quality pulse wave measurements, with signal degradation observed under loose placement conditions.

The proposed magnetic-based sensor represents a promising alternative to conventional pulse monitoring technologies. It combines high sensitivity, low hysteresis, and contact force measurement capabilities in a compact, wearable design. These features address well-known limitations in pulse wave monitoring, particularly standardization issues in signal acquisition related to sensor positioning and contact pressure variations. Future work will explore applications in diverse populations with different vascular conditions and assess performance during motion.

The increasing digitalization of work models and the widespread adoption of remote and hybrid work have profoundly transformed workers’ habits, leading to greater sedentary behavior and prolonged exposure to video terminals. This evolution has made the adoption of physiological monitoring strategies increasingly urgent to prevent conditions of stress, fatigue, and related pathologies. In this context, an alternative approach involves the development of nearable devices, namely solutions integrated into everyday objects that enable continuous physiological monitoring without compromising user comfort and freedom of movement. Smart systems such as chairs, cushions, and mattresses thus represent a non-invasive alternative to traditional wearable devices.

This work proposes the development of an innovative nearable device based on Fiber Bragg Grating (FBG) technology, integrated into a structure fabricated via 3D printing through Fused Deposition Modelling (FDM) using thermoplastic polyurethane (TPU), and designed for the continuous monitoring of respiratory rate (RR) and heart rate (HR) in video terminal workers. The device consists of a sensing element composed of a circular base (40 mm diameter, 2 mm thickness), on which a dome-shaped extrusion (20 mm diameter, 2 mm height) was realized, specifically designed to optimize the transmission of mechanical deformations generated by the expansion and contraction of the rib cage during the respiratory cycle to the FBG sensor (10 mm length, λ_B = 1533.06 nm, 96.83% reflectivity, acrylate recoating). The fiber integration into the structure was achieved during the printing process through a controlled print-pause procedure, ensuring correct positioning and preserving the optical and mechanical properties of the system.

Following fabrication, the device underwent metrological characterization to evaluate its strain sensitivity (S_ε). Controlled compression tests were performed, simulating thoracic load with up to 5% deformation applied to the dome of the sensing element. Data analysis revealed an S_ε value of 0.028 nm/%. Subsequently, the system was preliminarily validated on two healthy volunteers subjected to controlled breathing protocols (quiet breathing and fast breathing). The signals acquired through an optical interrogator were processed using custom-developed algorithms in the MATLAB environment, enabling the extraction of physiological parameters through spectral analysis.

Experimental results demonstrated high accuracy in estimating RR, with mean absolute errors lower than 0.74 breaths/min and percentage errors below 3.5% under both tested conditions. HR estimation was also satisfactory, although an increase in error was observed under fast breathing conditions, consistent with evidence reported in the literature for nearable devices based on FBG technology.

The developed device represents a promising technological solution for non-invasive and continuous physiological monitoring in work environments. Future developments will include the optimization of the sensing geometry, the integration of multi-sensor architectures, and validation on a larger sample of subjects to consolidate the system’s effectiveness in real operational scenarios.

We conduct a multiscale study of a composite material comprising two solid phases undergoing a remodelling process of its microstructure. This remodelling is characterised by the evolution of inelastic distortions, described within the context of the Bilby-Kröner-Lee decomposition of the deformation gradient tensor [1] in a multiscale framework [2-4]. The study begins with the derivation of governing equations that capture the behaviour of the composite's constituents, rooted in purely mechanical setting. To bridge the gap between the evolution of the microstructure, dictated by the progression of the inelastic distortions, and macroscopic properties, we employ the two-scale Asymptotic Homogenisation (AH) technique [2-4]. This mathematical framework enables the upscaling of fine-scale inelastic distortions—responsible for variations in the elasticity moduli of the composite—to coarse-scale descriptions, as well as the geometrical characteristics of the composite's internal structure. Our ultimate objective is to address computational challenges associated with predicting the effective mechanical properties of remodelling composites [2]. These challenges arise due to the complex interactions between cell-level problems and homogenised macro-scale behaviours. By leveraging our approach, we derive analytical expressions for the effective coefficients, parameterised spatially and temporally by the evolving tensor of anelastic distortions. Particular attention is given to cases involving multilayered systems [3] or uniaxially fiber-reinforced composites [4].

References:
[1] Micunovic M., Thermomechanics of Viscoplasticity - Fundamentals and Applications. Springer, Heidelberg, Germany, 2009.
[2] Ramírez-Torres A., Di Stefano S., Grillo A., Rodríguez-Ramos R., Merodio J., Penta R., An asymptotic homogenization approach to the microstructural evolution of heterogeneous media. International Journal of Non-Linear Mechanics, 106:245–257, 2018.
[3] Giammarini A., Ramírez-Torres A., Grillo A., Effective elasto-(visco)plastic coefficients of a biphasic composite material with scale-dependent size effects. Mathematical Methods in the Applied Sciences, 48:926–979, 2025.
[4] Ramírez-Torres A., Roque-Piedra A., Giammarini A., Grillo A., Rodríguez-Ramos R., Analytical expressions for the effective coefficients of fibre-reinforced composite materials under the influence of inelastic distortions. Journal of Applied Mathematics and Mechanics (ZAMM), Accepted, 2025.

Acknowledgements:

A. Ramírez-Torres is supported by the Engineering and Physical Sciences Research Council [grant number EP/Y001583/1]. This grant is funded by the International Science Partnerships Fund (ISPF) and the UK Research and Innovation (UKRI). A. Grillo is partially supported by MIUR (Italian Ministry of Education, University and Research) through the PRIN project n. 2017KL4EF3 on “Mathematics of active materials: From mechanobiology to smart devices”, the PRIN project n. 2020F3NCPX on “Mathematics for industry 4.0 (Math4I4)”. A. Grillo's research group is funded by the European Union ‐ Next Generation EU. A. Grillo and A. Giammarini have been supported by the Research Project Prin2022 PNRR of National Relevance P2022KHFNB on “Innovative multi‐scale approaches, possibly based on Fractional Calculus, for the effective constitutive modeling of cell mechanics, engineered tissues, and metamaterials in Biomedicine and related fields” granted by the Italian MUR. This study was carried out within the MICS (Made in Italy – Circular and Sustainable) Extended Partnership and A. Giammarini received funding from the European Union Next-GenerationEU (PIANO NAZIONALE DI RIPRESA E RESILIENZA (PNRR) – MISSIONE 4 COMPONENTE 2, INVESTIMENTO 1.3 – D.D. 1551.11-10-2022, PE00000004). This manuscript reflects only the authors’ views and opinions, neither the European Union nor the European Commission can be considered responsible for them. R. Rodríguez‐Ramos thanks to Chamada CNPq 09/2023 PQ‐2 Productividade em Pesquisa, processo . 307188/2023‐0 e ao Edital UFF PROPPI . 05/2022.

In the context of cell mechanics, we depict a multi-scale continuum approach to study the exchange of mechnanical actions between focal adhesions (FAs), the extra-cellular matrix (ECM) and living cells [1]. In particular, we investigate the role of force- and stress-type stimuli in determining the remodelling, or structural time evolution, of both FAs and ECM, and we investigate how possible heterogeneities attainable at the micro-scale, i.e., the scale characterising the internal structure of each component of focal adhesions, may influence their behaviour [2, 3, 5]. With reference to some structural models available in the literature, we describe a focal adhesion as a sandwich structure, accounting for three main components: the adhesion plaque, the integrins receptors and the ECM [1, 3]. Following previous works [1, 2, 3], we employ a mono-dimensional shear lag model [2, 3], so that, both the adhesion plaque and the ECM are modelled as linear elastic straight fibres subjected to axial deformation only, while the family of integrins receptors is represented by a system of elastic and non-elastic forces. Furthermore, we consider the description of focal adhesions’ dynamics by accounting for their remodelling and micro-scale inhomogeneities. To achieve this, we follow and adapt to our scopes some tools of non-linear elastoplasticity and we adhere to techniques of Asymptotic Homogenization to elucidate how the micro-structure of focal adhesions influences their overall behaviour [3, 4, 5]. In this regard, we obtain closed form of the effective coefficients characterising the FA-ECM complex, with the latter containing both constitutive and geometric information available at the micro-structure.

The obtained multi-scale model, and the related field equations, are solved and the forces exchanged by focal adhesions with the ECM and cells are studied. Detailed findings from this investigation are summarized in [5].

References

[1] Cao X., et al., “A chemomechanical model of matrix and nuclear rigidity regulation of focal

adhesion size,” Biophys. J., 109.9, 1807-1817 (2015).

[2] Di Stefano S., et al., “On the role of elasticity in focal adhesion stability within the passive

regime,” Int J Non Linear Mech, 146, 104157 (2022).

[3] Di Stefano S., et al., “On the role of friction and remodelling in cell–matrix interactions: a

continuum mechanical model,” Int J Non Linear Mech, 142, 103966 (2022).

[4] Ramírez-Torres-Torres A., et al., “An asymptotic homogenization approach to the microstructural evolution of heterogeneous media,” Int J Non Linear Mech, 106, 245-257 (2018).

[5] Di Stefano S., et al., ''Homogenised structural behaviour of remodelling cell-matrix systems: the case of focal adhesions''. In preparation.

Within the framework of "Hybrid Mixture Theory" [1], multicellular aggregates can be formalized as biphasic continuum media featuring a fluid phase, typically identified with an interstitial fluid carrying nutrients, and a solid phase representing, e.g., cells, protein filaments, extracellular matrix and other biological components.

Experimental evidence [2] shows that, under external loading, the solid phase of multicellular aggregates can undergo irreversible deformations, so that, after relaxation, these biological structures tend not to recover their shape prior to the application of the load [3]. Some authors [4,5] have proposed to attribute this behavior to internal structural reorganizations, which, in some degree, share similarities with the inelastic processes occurring in non-biological media. While these structural transformations belong to the class of phenomena named “remodeling” in the biomechanics community, their presumed resemblance with inelasticity has suggested, among other possibilities, to describe them by multiplicatively decomposing the deformation gradient tensor of a given multicellular aggregate into an elastic and an inelastic (remodeling) part (see, e.g., [3] and the references therein). On the other hand, also the fluid phase contributes to the overall dissipative behavior of the systems under study, for example through its exchange interactions with the solid phase. In the literature, the inelastic aspects of remodeling and the dissipation introduced by the fluid are usually addressed by resorting to models that are of grade zero in the distortions and that rely on flow laws of Darcian type for the fluid. In spite of their utility, however, these models are unable to resolve explicitly the interactions between the multicellular aggregates and the surfaces of the apparatuses with which they are in contact during experiments. An example of these interactions could be the change in ductility localized near the contact regions of a specimen with the testing device.

To cope with this lack of information, an interesting, yet relatively unexplored, research direction could be the introduction of higher-order descriptors both for the inelastic distortions and for the fluid kinematics. In particular, the flow of the interstitial fluid may be characterized by a non-negligible viscosity, thereby undermining the hypothesis of Darcian-like regimes. In this case, one possibility is to account for the so-called Brinkman correction in the expression of the overall fluid stress tensor [6].

In this presentation, we take a step in this direction by extending the model of Gurtin&Anand [7] to the theory of biphasic mixtures, and coupling it with a Darcy–Brinkman model for the fluid flow. We follow the paradigm of the Principle of Virtual Power [8] to obtain the dynamic equations of the system, and we discuss some related aspects of configurational and analytical mechanics. Finally, we compare numerically some preliminary results [9] with some well established models taken from the literature.

References:

[1] Bennethum, L.S., et al.: Transport in Porous Media 39(2): 187–225 (2000).

[2] Marmottant, P. et al.: Proc Natl Acad Sci USA. 106(41): 17271-5 (2009).

[3] Di Stefano, S., Giammarini, A., Giverso, C., et al.: Z. Angew. Math. Phys. 73: 79 (2022).

[4] Giverso, C., Preziosi, L.: Math Med Bio 29: 181-204 (2012).

[5] Ambrosi, D., Preziosi, L.: Biomech Model Mechanobiol 8(5): 397-413 (2009).

[6] Brinkman, H. C.: Applied Scientific Research. 1(1): 27–34 (1949).

[7] Gurtin, M. E., Anand, L.: Int. J. Plast., 21(12): 2297-2318 (2005).

[8] Germain, P.: SIAM J. Appl. Math. 25(3): 556–575 (1973).

[9] Giammarini, A., Pastore, A., Ramìrez-Torres, A., Grillo, A.: To be submitted.

.

We present a phase field model describing cell diffusion in a soft tissue, with particular attention to their active behavior in sensing nearby cells and the mechanical properties of the surrounding environment to guide their movement. The model derivation is framed within the principles of power balance and micro-balance expenditure, and, in principle, it has a general applicability to a population of mesenchymal cells moving inside a visco-elastic environment and “activated” by several chemo-mechanical cues [1]

A central aspect of the model relates the derivation of the cell population chemical potential: it is made of by the superposition of the entropic energy - homogeneous convex free energy - and the active potential induced by the microforces, responsible for triggering the instability process through the spinodal decomposition dynamics. The active component of the chemical potential is adherent to the phenomenological law firstly proposed in [2]. Additionally, the microforce balance is enhanced by the power conjugate associated with variations in the concentration gradient, orienting the cell diffusion toward a stationary limit pattern during a metastable dynamics [1,3]. This vector field could potentially model any directional cue or bias characterizing the interaction between cells and the surrounding elastic environment, making the cell activity at the microscale to emerge on the tissue mechanics.

In testing the descriptivity and the predictivity of this active cell population dynamics model, we apply it yo the case study of the hepatic fibrosis: a pathological process characterized by excessive deposition of extracellular matrix proteins in the liver, typically as a response to chronic injury caused by factors such as viral infections, alcohol abuse, or metabolic disorders. Over time, this scarring can disrupt normal liver architecture and function, leading to abnormal stiffening of the tissue and impairment of blood perfusion. The myofibroblasts are the activated mesenchymal cells responsible for extracellular matrix production during the over mentioned process. The progression stages of the pathology characterize for regular patterns deposition within the hepatic lobule - the fundamental functional unit of hepatic tissue - which are successfully replicated by our FEM simulation, with appropriate constitutive choices of the active microforce and microcouple fields.

References

• [1] F. Recrosi, A. Tatone, G. Tomassetti, “Driving forces in cell migration and pattern formtion”, arXiv, Soft Condensed Matter, arXiv:2410.22273,(2024).
• [2] G.F. Oster, J.D. Murray, A.K. Harris,”Mechanical aspects of mesenchymal morphogenesis”, J. Embryol. exp. Morph. 78 83–125. (1983).
• [3] F. Recrosi, R. Repetto, A. Tatone, G. Tomassetti ”Mechanical Model of Fiber Morphogenesis in the Liver”, Proceedings of XXIV AIMETA Conference, Lecture Notes in Mechanical Engineering. Springer, (2019).

Introduction

Cell growth and proliferation reduces under mechanical loading and confinement [1], but the mechanisms underlying this behaviour are not understood. Cells are subject to a volume checkpoint that restricts proliferation of small cells [2], and recent evidence further suggests that dilution of key cell-cycle inhibitors may govern this behaviour. In this study, we propose a biophysical model that couples transcriptional-translational kinetics with the hydromechanics of cell growth to investigate the feedback between cycle inhibitors, biomolecule synthesis, and mechanical loading in regulating cell growth and division.

Methods

Cell fluid volume is controlled by a balance between hydrostatic pressure and osmotic pressure arising from cytoplasmic solutes [3]. We first developed a mathematical model to predict mRNA and protein production, linking free amino acid availability to translation. Synthesised biomolecules in turn increase osmotic pressure to facilitate cell cycle growth, with volume changes arising from water flux further influenced by membrane tension and external mechanical loading. Synthesis of charged proteins additionally influences ion transport, with feedback to cellular osmotic pressure. We implemented this model within PhysiCell, an open source agent-based modelling framework to analyse discrete cell-cell interactions. This framework was further coupled with a custom finite element solver to describe contact between cells and surrounding hydrogel and matrix.

Results and Discussion

Our analyses reveal that biomolecule synthesis during the cell cycle increases osmotic pressure to drive cell growth. Free amino acids, mRNA, and proteins govern cell volume by modulating both intracellular osmotic pressure and membrane potential. With cell growth, a key cycle inhibitor Whi5 is diluted in agreement with experimental observations [4]. Sufficient dilution facilitates G1/S phase progression, with our model characterising the volume dependency of yeast cell division. However, when cells experience sufficient external loading, our model predicts that growth reduces due to increased hydrostatic pressure and cell division is restricted by insufficient Whi5 dilution and macromolecular crowding. Simulations also suggest that smaller cell birth size leads to an increased G1 phase duration, as observed by Schmoller et al [4]. Extension to the agent-based framework revealed how compaction arises from cell-cell adhesion, amplifying growth-induced pressure experienced by proliferative cells. Simulations reveal that mechanical feedback between surrounding hydrogel deformation and contacting cells further suppresses fluid intake and the rate of cell growth, underpinned by insufficient Whi5 dilution and reduced biomolecule synthesis in agreement with experimental reports [1]. This framework will next be mapped to understanding stress-dependent tissue and tumour growth, where the protein Rb may play a similar role to Whi5 [5]. Thus, leveraging insight from suppression of yeast growth can shed light on the role of cycle inhibitors on tumour progression and identify potential therapeutic targets.

References

1. Alric et al. (2022), Nature Physics, 18.

2. Varsano et al. (2017), Cell Reports, 20

3. McEvoy et al. (2020), Nat Commun, 11.

4. Schmoller et al. (2015), Nature, 526.

5. Zatulovskiy et al. (2020), Science 369.

Heart morphogenesis is mechanically driven by inhomogeneous and anisotropic growth patterns, which involve buckling and twists of an initially tubular shape [1]. Genetic mutations and growth defects in this complex but robust process may yield aberrant shapes which may have detrimental consequences in life expectancy or even fatal effects. For instance, greb1l mutants lack sufficient growth in heart outflow track and are absent of torsional growth, which is believed to cause crisscross malformations [2].

Despite recent advances in microscopy data and determination of heart geometry during its morphogenesis, it is still difficult to predict and quantify the growth patterns that are responsible of tis final shape. This is due to the variability of the shape, and the presence of anisotropies. For this reason, computational techniques that can estimate the growth distribution in different embryos become extremely useful to determine differences and growth defects.

In this work, we present a methodology for computing growth distributions from a series of three-dimensional snapshots [3]. By resorting to a finite element discretization of our domain, and a piecewise constant distribution of orthotropic growth tensor, we formulate an inverse problem that aims at computing a set of optimal patterns that match the resulting shape in a least square sense. Additionally, we also include the unknown boundary conditions as part of our inverse problem, given the impossibility to accurate measure boundary tractions.

Due to the multiplicity of the solutions, the optimization is regularized and iteratively solved for reducing the effects of the regularization parameters. We demonstrate through a series of synthetic examples that the growth can be correctly reproduced, and also apply the methodology to a series of heart subdomains. We also show that the optimal conditions of the inverse problem have a Hamiltonian structure, which can be exploited for the numerical solution of the problem. We finally also show that the same problem can be employed to predict contractility patterns in organism locomotion [4].

REFERENCES

[1] Le Garrec et al. A predictive model of asymmetric morphogenesis from 3D reconstructions of mouse heart looping dynamics. eLife 2017;6:e28951. DOI: https://doi.org/10.7554/eLife.28951

[2] S Bernheim, A Borgel, JF Le Garrec, E Perthame, A Desgrange, C Michel, L Guillemot, S Sart, CN Baroud, W Krezel, F Raimondi, D Bonnet, S Zaffran, L Houyel, SM Meilhac. Identification of Greb1l as a genetic determinant of crisscross heart in mice showing torsion of the heart tube by shortage of progenitor cells., Dev Cell, 58(21), 2023.

[3] C Olivesi, JJ Muñoz. Inverse analysis for the computation of growth and boundary conditions in elastic bodies. Comp. Mechanics, 2024. doi.org/10.1007/s00466-024-02546-5

[4] A Bijalwan and JJ Muñoz. Adjoint-based optimal control of contractile elastic bodies. Application to limbless locomotion on frictional substrates. Comp. Meth. App. Mech. and Eng., 420:116697, 2024.

To form new sprouts during angiogenesis, endothelial cells must coordinate their migration through biophysical and biomechanical signaling between each other and the micro-environment. A relatively simple model of angiogenesis is intersegmental vessel (ISV) formation in zebrafish. While various molecular, cellular, and mechanical factors coordinate ISV pathfinding between the somites, the specific contributions of ECM components remain incompletely understood. Here we hypothesize that guidance through ECM molecules laid down in the intersomitic space confines a process of self-organized pattern formation to the intersomitic space. We combine theoretical and experimental approaches in the zebrafish to investigate this hypothesis. Firstly, we developed a hybrid mathematical model to study the effect of ECM mechanics on a self-organized mechanicsm of endothelial network formation. A network of fibers represents a generic ECM, which is simulated using a coarse-grained mass-spring system, while endothelial network formation is modeled with a Cellular Potts Model (CPM) and partial-differential equation model for intercellular signaling through small cytokines. This framework was extended to incorporate ISV-specific factors such as VEGF and semaphorin signalin, cell polarization, and integrin-based mechano-sensitive of adhesion of cells to the ECM. In absence of an ECM, our model predicts the formation of endothelial network-like structures, as in our previous work. However, in presence of an ECM, the sprouts of consisting of endothelial cells migrate along high concentrations of the ECM. Thus the model predicts that if the concentration of ECM were reduced in the intersomitic space, the endothelial cells should organize into network-like structures. To investigate the contributions of the ECM to ISV formation in the zebrafish, we employed morpholino-mediated gene knockdown of ECM components in endothelial cell- and ECM-tagged zebrafish lines, coupled with high-resolution laser scanning confocal microscopy. Simultaneous knock-down of one type of ECM protein delays ISV sprouting during the first six hours; however, vessel development eventually proceeds normally, albeit with reduced cell proliferation and minor hypersprouting events. Notably, simultaneous knockdown of an additional ECM protein resulted in aberrant vessel pathfinding, leading to disorganized, incomplete ISV development, resembling endothelial network patterns formed through self-organization by endothelial cells in in silico and in vitro. Furthermore, we observed that endothelial cells interact with and migrate along laminin- and fibronectin-rich pathways, as revealed by zebrafish reporter lines, further underscoring the role of these ECM components in guiding vessel formation. Our simulations suggest that ECM stiffness may significantly influence endothelial cell migration, with cells potentially traveling further along VEGF gradients on stiff ECM compared to soft ECM under moderate VEGF sensitivity, providing a potential explanation for the observation in ECM component knockdowns. These results indicate that ECM-regulated tip cell migration could be a key determinant of ISV growth speed and patterning. All in all, by integrating experimental and theoretical approaches, our study suggests that an interplay between endothelial self-organization and ECM-guided pathfinding contributes to ISV formation.

Zero-dimensional (0D) cardiovascular models are widely used to simulate hemodynamics across the entire circulation, supporting applications from clinical decision-making to surgical planning. However, their real power lies in solving inverse problems, where patient-specific parameters are estimated from measurable outputs such as pressures or flow rates. These tasks, parameter estimation and uncertainty quantification (UQ), are computationally intensive, often requiring thousands of model evaluations. In this work, we present a comprehensive metamodeling pipeline that enables fast and reliable inverse problem-solving for 0D cardiovascular systems.

We compare three surrogate modeling approaches—feed-forward neural networks (NNs), polynomial chaos expansion (PCE), and Gaussian processes (GPs)—in their ability to emulate 0D models and support inverse tasks. Surrogates are trained on synthetic datasets generated via Saltelli’s sampling scheme and evaluated across three representative 0D models: (1) portal pressure prediction after liver resection, (2) hemodynamic modeling of pulmonary arterial hypertension (PAH) before and after Potts shunt placement, and (3) contrast-agent transport for perfusion assessment. These cases cover both scalar and time-series outputs, with the latter addressed using LSTM architectures.

Focusing on the PAH model, we demonstrate the full inverse pipeline using a trained NN surrogate. First, we perform variance-based sensitivity analysis using Monte Carlo Sobol indices to identify the most influential parameters. These insights guide the parameter selection and improve interpretability. Next, we solve the inverse problem—recovering unknown model parameters from observed outputs—using gradient-based optimization with automatic differentiation. By reparametrizing inputs through bounded transformations, we ensure physiological plausibility throughout the estimation process. This approach reliably identifies key parameters such as vascular resistances, chamber properties, and shunt characteristics, with typical convergence times under two minutes.

To quantify uncertainty in the inverse solution, we propagate input data noise through the inverse problem via a Monte Carlo approach. For each sampled clinical measurement set, we solve the parameter estimation problem using the surrogate, then propagate the resulting parameter distribution through the model with the shunt in place. This yields output distributions (e.g., for pressures, stroke volumes, flow ratios) that inform confidence in predicted outcomes.

Compared to PCE and GP, neural networks emerge as the most robust choice. They support fast training on large datasets, integrate naturally with automatic differentiation, and scale efficiently to high-dimensional input spaces. This makes them ideal candidates for real-time or near-real-time clinical inference tasks.

In conclusion, we propose a surrogate-enabled framework tailored to inverse problems in cardiovascular modeling, enabling rapid and personalized parameter estimation and UQ. This approach paves the way toward computationally efficient digital twins for clinical use.

Numerical cardiovascular modeling is a growing tool for clinical applications aimed at personalized medicine. As such, lumped parameter models offer computational efficiency, yet, require calibration with often sparse clinical data. This study [1] compares two established cardiac chamber models—Time-Varying Elastance (TVE) and Single-Fiber (SF) models —through sensitivity and identifiability analyses to assess their suitability for patient-specific applications. The case of a young pulmonary arterial hypertension (PAH) patient serves as the clinical context, although the methodology is applicable to other conditions.

The TVE and SF models were integrated into a lumped parameter closed-loop circulation capable of simulating whole-body hemodynamics. Patient-specific data from a 13-year-old with PAH were used to calibrate the models through inverse problem optimization employing the CMA-ES method. After successful tuning of the two models, sensitivity analysis was conducted in order to quantify the impact of input parameters on clinically relevant outputs, such as ventricular pressures and volumes, based on the total Sobol indices. Physiological constraints were enforced to ensure the outputs remained within clinically relevant bounds, while an extensive literature review had to be performed to define the input ranges. The identifiability of the sensitive parameter set can be then assessed using Profile Likelihood analysis (PLA). It is a step-wise process, which involves inverse problem solving (similarly to the calibration step) and evaluates whether a unique set of model parameters can be reliably estimated from the available clinical data.

The SF model demonstrated superior performance in reproducing patient-specific hemodynamic data, accurately capturing nonlinear ventricular pressure-volume dynamics and key parameters such as stroke volume and pressure levels. Sensitivity analysis identified dominant parameters affecting outputs, which showed that there is a significant interaction between parameters describing the systemic circulation and pulmonary hemodynamics, and vice versa. This highlights the importance of studying the whole circulation, particularly in diseases traditionally assumed to affect only one side. Identifiability analysis revealed that the SF model’s parameters were more reliably estimable than those of the TVE model, which showed limitations in the identification of key parameters such as ventricular and atrial elastances.

Despite the fact that the TVE model offers simplicity and computational efficiency (due to its linear nature), our comparative analysis in the setting of pulmonary hypertension indicates that the SF model is more suitable for personalized cardiac simulations. The limited physiological interpretability of the TVE model not only required more clinical data to find suitable personalized parameters but also made the determination of input ranges for the sensitivity analysis significantly more challenging. The SF model’s detailed representation facilitates better alignment with clinical data, which is essential for personalized medicine applications. Our study underscores the importance of cardiac modeling choice based on disease case study and data availability. This study also highlights the importance of comprehensive data collection, sensitivity analysis and model validation in advancing personalized medicine. As a future step, it would be insightful to compare non-linear extensions of the TVE model to the SF one.
[1] Haghebaert et al., J of Physiology, accepted

Patient-specific computational models of the cardiovascular system can inform clinical decision-making by providing physics-based, non-invasive calculations of quantities that can not be measured or are impractical to measure and by predicting physiological changes due to interventions.

In particular, mixed-dimensional 3D-0D coupled models can represent spatially resolved 3D myocardial tissue mechanics and 0D pressure-flow relationships in heart valves and vascular system compartments, while accounting for their interactions in a closed-loop setting. They require significantly less computational effort than fully spatially resolved 3D fluid dynamics modeling, which is not required for the considered clinical application.

We present an inverse analysis framework for the automated identification of a set of 3D and 0D patient-specific parameters based on flow, pressure, and cine cardiac MRI measurements. We propose a novel decomposition of the underlying large, nonlinear, and mixed-dimensional inverse problem into an equivalent set of independently solvable, computationally efficient, and well-posed inverse subproblems. This decomposition is enabled by the availability of measurement data of the coupling quantities and ensures a faster convergence towards a unique minimum. For example, if measurements of the right ventricular and pulmonary arterial pressures and pulmonary flow are available, then the parameters of the pulmonary valve submodel can be identified in a completely decoupled manner from neighboring submodels. It avoids the propagation of identification errors, given that submodels do not rely on computed coupling quantities from neighboring submodels to evaluate their response. With the proposed decomposition, we formulate simple and interpretable objective functions and avoid the convergence of the inverse problem to local minima, which do not satisfy all required similarities between computed and measured target quantities. The inverse subproblems are solved with a L-BFGS optimization algorithm and an adjoint gradient evaluation.

The proposed framework is demonstrated in a clinical case study of an adult repaired Tetralogy of Fallot (ToF) patient with severe pulmonary regurgitation. The identified parameters provide a good agreement between measured and computed flows, pressures, and chamber volumes, ensuring a patient-specific model response. The outcome prediction of an in silico pulmonary valve replacement using the personalized model is physiologically consistent and correlates well with postoperative measurements. To successfully implement the proposed approach, detecting, correcting, or removing inaccurate or inconsistent measurements with the guidance of experienced clinicians is required.

The proposed framework is essential for developing accurate and reliable cardiovascular digital twins and exploiting their predictive capabilities for intervention planning.

Our bones are impressive structures that constantly adapt to habitual mechanical load through a dynamic process known as bone remodelling. This continuous renewal allows bones to maintain their mechanical stability. Nevertheless, excessive overload can disrupt the delicate balance of bone remodelling leading to the formation of microcracks. However, as long as these microdamages remain within a certain threshold, they even trigger an enhanced remodelling response in these areas, allowing the bones to heal these microcracks in a targeted manner [1].

Recent research [2] has shown that the so-called flexoelectric effect—the ability of bone material to generate an electric potential under inhomogeneous deformation—plays a crucial role in coordinating the cellular mechanisms essential for the healing of microcracks. A deeper understanding of these regenerative processes, combined with their modelling and simulation, is therefore essential for developing more effective medical treatments and therapeutic strategies.

We propose a novel methodology that phenomenologically accounts for the trabecular microstructure of bone and size-dependent material effects using a micromorphic framework [3]. This framework has been extended to include nonlinear electro-elastic and flexoelectric energy contributions within the constitutive equations. To evaluate the approach, we use a cracked cantilever beam as a representative model, approximating a bone sample with a microcrack—such as those that may occur in the femoral neck region—and commonly employed in experimental studies [2]. In the presence of a bone crack, the flexoelectric effect causes an electric field to be generated around the crack when asymmetric deformation (e.g., bending) is applied. The stress gradient is highest at the tip of the crack, resulting in the greatest electric field. The effect is therefore most pronounced at the crack tip and decreases with further distance. The generated electric field acts as a biological trigger, inducing osteocyte apoptosis, which is followed by the bone-building activities of the osteoblasts [1,2]. The activation of these cellular mechanisms is captured in our model by an increase in nominal bone density in the vicinity of the crack tip. The presented analysis investigates the influence of the key model parameters on the magnitude of the electric potential and, consequently, on the nominal bone density. Additionally, we present an enhanced approach that specifically analyses the critical range of electric fields between 1 and 10 kV/m, which induce programmed osteocyte cell death (apoptosis), and examines their impact on the growth of nominal bone density in these regions.

[1] Heino T, Kurata K, Higaki H, Väänänen K. 2009. Evidence for the role of osteocytes in the initiation of targeted remodeling. Technology and Health Care. 17:49–56
[2] Núñez Toldrà R, Vasquez-Sancho F, Barroca N, Catalan G. 2020. Investigation of the cellular response to bone fractures: Evidence for flexoelectricity. Scientific Reports. 10:254
[3] Titlbach A, Papastavrou A, McBride A, Steinmann P. 2023. A novel micromorphic approach captures non-locality in continuum bone remodelling. Computer Methods in Biomechanics and Biomedical Engineering. 27(8):1042–1055

Recent advancements in manufacturing technologies have enabled the development of microstructurally architected materials, with functionally graded lattices emerging as promising candidates for biomedical applications. Their ability to replicate the complex structure of native bone makes them suitable for orthopedic and dental implants. The performance of these materials depends on both biomechanical compatibility and biological integration. Accordingly, recent studies have explored the mechanical behavior of TPMS structures within manufacturability constraints [1], and employed density-based topology optimization on homogenized lattices to design patient-specific hip implants [2]. This study aims to define a more accurate mechanical optimization approach to mitigate the stress shielding effect that may occur in the prosthesis-bone system due to differences in compliance between the bone and implant materials. The stress shielding effect can progressively compromise the implant's load-bearing capacity, as the surrounding bone progressively thins over time, thereby negatively affecting the implant's durability and functionality. The developed methodology is based on a dataset obtained from micro-Computed Tomography (μCT) scans of a series of human hip bone samples. From this dataset, specific morphological and mechanical characteristics have been extracted. In particular, the volume fraction is calculated as the ratio between the dimensions of the bone tissue islands and the total sample volume, while the trabecular thickness is estimated through direct interpolation by superimposing ellipsoidal domains onto the segmented bone tissue. The anisotropy is evaluated using the Mean Intercepts Length (MIL) method. The statistical behavior of these geometrical features is incorporated into a computational study to produce a continuum equivalent of the resected bone prior to implantation. Given the high computational costs related to stochastic computational homogenization, it is demonstrated that the complex bone material with random microstructure can be substituted with a simplified ergodic model. Several numerical homogenization techniques are employed to represent composites with random structures: from stochastic collocation methods to polynomial chaos, as well as Monte Carlo methods and Stochastic Finite Elements (SFEM). The latter are the most commonly used for such composites [3,4]. The homogenized continuum equivalent allows for the evaluation of the minimal achievable stress discontinuity at the prosthesis-bone interface, with accuracy governed by the level of geometric detail incorporated in the model. The stress jump is quantified through finite element analysis of the implant under various loading conditions. Future developments of this work will involve the formulation of an inverse problem, where the homogenized continuum serves as the mechanical target for a density-based topology optimization of a graded, periodic lattice structure—such as TPMS.

References:

1. T. Poltue, et al, IJES2 211,106762, 2021

2. P. Muller et al, Nature/Scientific Reports 14:5719, 2024

3. M. Ostoja-Starzewski, Int. J. Solid Struct., 5, 19, 2429-2455, 1998.

4. P. Steinmann et al, Comput. Methods Appl. Mech. Engrg. 357, 112563, 2019.

Mechanics-based in silico models of the human brain are gaining importance in understanding the role of mechanics in physiological and pathological processes such as brain folding, brain injury, ageing, and age-related diseases. Those models rely on experimental data for multiple purposes. The development or choice of appropriate models requires an understanding of underlying mechanisms that relies on observations made in experiments. Once a model is chosen, data are required for parameter identification and validation. One major challenge for further improvements of existing models is the limited access to data. Due to the vulnerability of the brain, in vivo testing is restricted to small strains. Human tissue from body donations is not always available and the quick degradation of brain tissue leads to a short time frame for post mortem data acquisition. As a response to the limited data availability, recent studies have attempted to gain insights into the mechanics of brain tissue through purely computational methods. Examples are micromechanical material models of white matter that are based on the assumption that aligned axons cause transversely isotropic material behavior. Here, we present results from ex vivo mechanical testing and histological analysis of human brain white matter and brain stem. Histological analyses of mechanically tested samples suggest alternative explanations for the direction-dependent behavior of tissue from the corpus callosum: in this brain region, not only axons and glial cells, but also blood vessels, which are thicker and stiffer than axons, are highly aligned. We further show that, while axons in the brain stem may appear unidirectional in diffusion tensor imaging data of lower resolution, the brain stem does not show transversely isotropic material behavior. This aligns with the complex anatomy of the brain stem, which not only contains axons of varying directions but also gray matter. With respect to brain modeling, our results suggest that continuum-mechanics based models combined with in vivo data and detailed anatomical knowledge may be a more robust, flexible, and cost-effective approach than those that explicitly model microstructural tissue components such as axons and cells. More generally, our results show that it is necessary to gain a more thorough understanding of brain tissue mechanics before focusing on a single modeling approach. A close collaboration with anatomists, clinicians, and biologists is beneficial to ensure that imaging data are interpreted correctly. In conclusion, while in vivo data are the best choice to be fed into ready-for-application, patient-specific models, a combination of in vivo and ex vivo data is needed to first elucidate underlying mechanisms and choose appropriate modeling approaches.

The brain is recognized as one of the most intricate organs in the human body, and modeling its mechanical behavior has posed significant challenges for researchers over many years. Understanding the response of this multiphase tissue to mechanical loading and the reliable calibration of material parameters remains limited. Previous viscoelastic models utilizing two relaxation times have effectively captured aspects of brain tissue response; however, the Theory of Porous Media offers a continuum mechanical perspective to investigate the fundamental physical mechanisms, particularly the interactions between the solid matrix and the freely flowing interstitial fluid. Leveraging our established experimental testing protocol, this study conducts finite element simulations of cyclic compression-tension loading and relaxation experiments on human visual cortex (grey matter) and corona radiata (white matter) specimens. The findings indicate that solid volumetric stress is a critical element influencing the biphasic behavior of the tissue, as it significantly interacts with the porous effects governed by the permeability. Through inverse parameter identification, it is revealed that poroelasticity alone fails to adequately capture the time-dependent material behavior, particularly the significant hysteresis observed during cyclic loading. Incorporating viscous effects to introduce a second time-dependent mechanism greatly enhances model accuracy, facilitating satisfactory representations for both loading modes simultaneously. Our analysis illustrates that the first Lamé parameter is instrumental in influencing fluid flow and porous dissipation: lower values facilitate volumetric deformation and augment permeability effects, whereas higher values inhibit fluid flow and reduce poroelastic behavior. Additionally, the first Lamé parameter governs the temporal progression of deformation within the biphasic material in conjunction with the permeability. The insights provided shed light on the distinct contributions of viscous and porous effects. Nonetheless, the strong interdependencies among porous, viscous, and volumetric effects underscore the necessity for further experimental investigations to reliably ascertain all material parameters. Thereby, our findings stress the importance of calibrating parameter models in accordance with expected deformation and fluid flow regimes.

Rectal prolapse is a condition in which part of the rectum protrudes through the anus, significantly affecting patients' quality of life. This study aims to characterize the in vivo biomechanical properties of the posterior rectal wall using inverse finite element analysis (FEA). Two patient groups are included: one with rectal prolapse with gel and the other with pelvic prolapse without gel (control group). The goal is to quantify these properties through a minimally invasive approach, utilizing supero-inferior displacement measurements from MRI images acquired in the sagittal plane during the defecatory maneuver.

The study begins by simulating rectal prolapse using finite element analysis, modeling the stages of prolapse with properties from the literature to characterize the rectum, through hyperelastic models, and applying pressure that simulates the defecatory maneuver. These simulations provide insight into how the rectal wall deforms under prolapse conditions.

The supero-inferior displacements obtained from MRI scans are used as input in the inverse finite element analysis to optimize the mechanical parameters of the hyperelastic models. This process allows for the derivation of nonlinear and anisotropic properties of the rectal wall tissues, providing more accurate data on the resilience, elasticity, and deformability of the rectal wall.

By comparing the MRI data with simulation results, this study aims to optimize the in vivo biomechanical properties. The ultimate goal is to improve the understanding of rectal prolapse mechanics and develop personalized treatment strategies, particularly for surgical interventions.

In conclusion, this study integrates numerical simulations with clinical MRI measurements to better understand the biomechanical properties of the rectal wall in rectal prolapse patients. The findings are expected to contribute to more effective therapeutic interventions for prolapse conditions.

The female perineum plays a crucial biomechanical role during childbirth, undergoing significant stretching that frequently results in tissue trauma. Despite improvements in obstetric care, over 90% of vaginal deliveries are still associated with perineal injuries, underscoring the need for better predictive tools. Computational models offer a non-invasive approach to analyze the complex mechanics of childbirth and inform clinical interventions. However, the predictive accuracy of these models depends on the availability of reliable, physiologically accurate tissue properties, which remain poorly characterized for the female perineum. This study aims to address this gap by calibrating a visco-hyperelastic constitutive model to capture the mechanical behavior of the female human perineal body. To this end, the material parameters of the Holzapfel-Gasser-Ogden model were calibrated, as it effectively represents both nonlinear hyperelasticity and anisotropic behavior.

Perineal tissue samples, specifically from the perineal body, were obtained post-mortem from six women, pre-cooled, and then stored frozen. A cryostat was used to section each sample into 0.6 mm slices, which were then cut into a dog-bone shape. To prevent slippage during testing, custom-designed 3D-printed clamps were used. The specimens were subjected to a series of uniaxial mechanical tests, including stress-relaxation and ultimate tensile tests. Experimental data revealed key mechanical characteristics: stress-relaxation tests demonstrated a 40% reduction in stress over a 900-second interval, confirming the viscoelastic nature of the tissue, while tensile tests revealed non-linear stress-stretch behavior, with an average ultimate Cauchy stress of 224.70 ± 131.69 N and a stretch ratio of 1.97. Histological examination confirmed that the fiber orientation aligned with the direction of loading, supporting the choice of a transversely isotropic constitutive model.

To calibrate the material parameters, finite element simulations were performed using a dog-bone shape geometry and boundary conditions matching the experimental tests. A genetic algorithm was used to iteratively minimize the difference between experimental data and simulation output through a least squares evaluation function. To calibrate the visco-hyperelastic material, the viscous properties of the Generalized Maxwell model were initially determined using the mean normalized relaxation curve obtained from the relaxation part of the experiments. Once the viscous parameters were successfully determined, the next step involved calibrating the hyperelastic properties. This was done by comparing the model’s predictions with the absolute values obtained from the Cauchy stress vs stretch experimental data. Three distinct sets of parameters were calibrated, capturing the minimum, mean, and maximum mechanical responses observed across the donor cohort.

The calibration process demonstrated good agreement between numerical predictions and experimental results, validating the robustness of the chosen model and optimization strategy. Incorporating these material properties into finite element simulations would enable accurate simulation of physiological phenomena involving the perineal body. In the context of childbirth, such simulations could play a crucial role in identifying women at greater risk of perineal trauma and guiding personalized obstetric care. Overall, this study bridges experimental biomechanics and computational modeling, contributing to improve maternal health outcomes.

Traditional metallic orthopedic screws, despite their high mechanical strength and biocompatibility, often require a secondary surgical removal procedure. Bioabsorbable materials represent a compelling alternative, as they provide the necessary mechanical support during bone healing and progressively degrade in vivo. Among these, Mg-based alloys represent a promising combination of biocompatibility, cytocompatibility, biodegradability, and mechanical properties suitable for orthopedic applications. However, a major challenge of bioabsorbable medical devices lies in the controllability and predictability of their degradation rate and associated mechanical response. In such a scenario, advanced computational modeling represents an essential tool for optimizing the design and tailoring their performance in specific clinical settings. Screw degradation in biological tissue entails the solution of a moving interface problem evolving in a space-time domain. Traditional numerical methods, such as the Finite Element Method (FEM), struggle to handle these problems due to the necessity of dynamically updating the non-smooth contact surface in a complex living environment. Accordingly, alternative approaches, e.g., the phase field formulation, have been proposed but are still limited to specific applications and require a local identification of the contact boundary conditions. In the present contribution, we consider a nonlocal peridynamic formulation naturally enclosing the global solution of multi-field and multi-scale interfaces, circumventing the need to specify local contact boundary conditions and requiring ad hoc numerical procedures. In particular, we adopt and optimize the PeriFast/Corrosion scheme [1], originally developed under the assumption of constant corrosion currents, and generalize the numerical routine to handle time-varying currents proportional to the volume loss rate. We consider as a case study an M2 orthopedic screw made of Mg-10Gd alloy and submerged in a generic biologically-like electrolyte. Model tuning was performed by comparing and contrasting the computed volumetric loss with experimental results [2]. Extensive numerical analyses demonstrated the capability of the proposed PD model to capture the overall trend of the degradation process effectively. Specifically, the degraded screw geometry reveals that the corrosive attack initiates on all the wetted surfaces but affects the threading more significantly, as it is the area most exposed to the electrolyte. The study also analyses the mechanical response of the degraded screw at different time steps under axial and bending loads, thus identifying the critical conditions for orthopedics sustainability. Future studies will focus on coupling the proposed PD degradation model with approaches that simulate the simultaneous bone growth and remodeling process. This will allow to reproduce, among other aspects, individual biochemical pathways, paving the way for a patient-specific customization of these devices.

[1] L. Wang et al., «PeriFast/Corrosion: A 3D Pseudospectral Peridynamic MATLAB Code for Corrosion,» Journal of Peridynamics and Nonlocal Modeling, 2023.

[2] A. Hermann et al., «Combining peridynamic and finite element simulations to capture the corrosion of degradable bone implants and to predict their residual strength,» International Journal of Mechanical Sciences, 2022.

Myoelectric prostheses are still far from fully replacing the complete functionality of the lost limbs. Particularly challenging are the situations in which joint stiffness must be varied during task execution. The human ability to dynamically adjust the mechanical properties of the joints in response to environmental demands is fundamental for a natural interaction with the surroundings. The integration of this feature in upper-limb prostheses is currently under studies. Most of the approaches proposed in literature are focused on developing mechatronic devices with variable stiffness. Whereas, the possibility of changing prosthetic joint impedance via software could enhance the integration of the approach into various commercial devices.
EMG-driven musculoskeletal models are promising for estimating joint stiffness due to their high accuracy and ability to extract additional parameters, such as joint torques and angles. They typically assume Hill-type muscle behavior and use EMG signals to estimate muscle force. However, they suffer of two main drawbacks: 1) the computational cost of the most accurate approaches, such as Opensim models, often exceed the response time required for online prosthesis control (< 150ms). 2) these models require calibration datasets with EMG signals and force measurements, typically obtained through subject specific experimental setups, limiting calibration process due to the need for specialized instrumentation.
The contribution of this work is twofold. First, we implement an innovative musculoskeletal model calibration pipeline using kinematic data instead of torque and force information, in order to reduce experimental setup complexity. Second, for elbow joint stiffness estimation, we use the MuJoCo environment instead of OpenSim, since it offer same accuracy and allows much faster simulations, enabling real-time estimation.
The proposed calibration pipeline gives in input to MuJoCo model the muscles activation level and compares its output (i.e. elbow angle) with data obtained from an optoelectronic system. The resulting Root Mean Square Error (RMSE) is minimized through a Powell optimization process, which adjusts the parameters of the muscle model to better replicate the reference data. This loop continues until the simulated output closely matches the real movement. To validate the proposed approach, experiments were conducted in order to analyse how stiffness varies across different loads and elbow-shoulder configurations. Ten participants (aged 29 ± 3) performed static acquisitions in various shoulder (0, 90°) and elbow configurations (45, 90, 135°), as well as under different loading conditions (0, 1.5, 3.0 kg). For each combination of shoulder-elbow-load configuration, the task was repeated 3 times.
Results show that the estimated stiffness values fall within the range reported in the literature (0–40 Nm/rad). Furthermore, statistical and biomechanical analyses were conducted to assess the efficacy of the proposed approach. The results demonstrate that the calibrated musculoskeletal model successfully distinguishes stiffness variations across different joint configurations and loading conditions.
In conclusion, a new methodological approach is proposed to integrate the musculoskeletal model information into prostheses control. This will allow to have a more intuitive prosthesis control.

Model explainability is becoming crucial for machine learning (ML) to unveil the reasoning behind the system’s decision and justification of its response. In health domains, these aspects are relevant as they can provide arguments to the evidence-based medicine and can become determinant when providing decisions on patients management and care after discharge.

Neurological diseases, such as stroke, although extremely variable under etiological and pathological perspectives, are characterised by clinical characteristics of often severe motor and cognitive impairments affecting everyday life activities. During the disease progression, the prediction of motor recovery may contribute to identify timely therapeutic decisions.

The European Medicine Agency has recently underlined the potential benefits of digital health technologies to improve the quality of the care standards, in particular adopting wearable devices, telemedicine and robotics in medical applications.

Artificial Intelligence and ML techniques may play a significant role in interpreting data coming from these applications in order to provide reliable tools to support clinicians' decisions.

The aim of the PREDICTOR project, funded under the PRIN national programme, is to develop and validate a proof-of-concept of an innovative technological framework based on the integration of robotics and wearable sensors for the prediction of motor recovery in the early subacute phase of stroke survivors, who will undergo an upper limb robot-assisted rehabilitation program, by means of explainable ML techniques.

Data from a) a planar end-effector robotic device for upper limb rehabilitation, b) wearable sensors (for home monitoring), and c) clinical scales will be analysed and used to train specific ML algorithms whose final expected output is represented by the prediction of 1) upper limb motor recovery in terms of scores of clinical outcome scales and 2) parameters (such as viscosity, weight and stiffness) in order to improve the patient-robot interaction. The former is a long-term prediction (e.g. prediction at 3/6 months after the admission), the latter is a short-term prediction (e.g. prediction for the subsequent rehabilitation session).

Stroke remains a leading cause of disability worldwide, often resulting in long-term motor impairments that necessitate intensive physical rehabilitation. In recent years, robotic systems have been increasingly integrated into rehabilitation programs, offering objective kinematic measurements and the ability to deliver consistent, high-intensity therapy. However, despite their promise, there remains a pressing need to better personalize treatment pathways and optimize resource allocation in clinical settings. This presentation introduces a predictive model developed to forecast the rehabilitation outcomes of post-stroke patients undergoing robot-assisted therapy, utilizing an integrated set of clinical, demographic, and robotic kinematic data.

Our model draws from a rich dataset comprising clinical scales (e.g., Fugl-Meyer Assessment, Modified Ashworth Scale), demographic variables (e.g., age, gender, time since stroke), and detailed kinematic parameters captured during robot-mediated therapy sessions. Using advanced machine learning techniques, we trained and validated the model to predict patients' functional gains over the course of rehabilitation, providing clinicians with valuable foresight into expected recovery trajectories.

Beyond prediction, the model also addresses a critical clinical need: dynamic personalization of rehabilitation exercises. By analyzing a patient's motor capabilities and real-time performance data, the system suggests optimal robot parameters that tailor the difficulty of exercises to each individual’s current ability, following previous choices of expert clinicians. This ensures that therapy remains appropriately challenging, promoting neuroplasticity and maximizing functional recovery.

The proposed approach offers several benefits to healthcare systems. First, by accurately predicting rehabilitation outcomes early in the therapy process, clinicians can make informed decisions about resource allocation, prioritizing intensive interventions for patients most likely to benefit. Second, the model's ability to recommend exercise parameters helps therapists fine-tune rehabilitation plans without relying solely on trial and error or subjective assessments, thereby saving time and enhancing treatment efficacy. Lastly, this personalized strategy fosters a patient-centered rehabilitation process, empowering patients through appropriately scaled challenges and enhancing engagement and motivation.

In a retrospective study across multiple clinical sites, our model achieved a good predictive accuracy and demonstrated robust generalizability across different patient profiles. Moreover, preliminary feedback from clinicians indicated that the model's parameter recommendations were both feasible and helpful in clinical decision-making.

This presentation will detail the architecture of the predictive model, the methodologies employed for feature selection and model training, and the validation results. Additionally, we will discuss case studies highlighting how model-driven parameter suggestions were accepted on average by the clinical personnel. Finally, we will address potential limitations, such as data quality challenges and the need for prospective validation, and outline future directions.

By providing a data-driven framework for both outcome prediction and exercise personalization, our work moves towards a future of smarter, more efficient, and more individualized rehabilitation for post-stroke patients.

There is growing interest in the application of mathematical and numerical models to gain deeper insight into physiologically interesting problems. For instance, cardiovascular disease of general interest to modellers and clinicians given the large global disease burden. Via mathematical models and simulation, we are able to gain a deeper insight into disease processes than is necessarily feasible or ethical to gain via physiological experimentation.

All mathematical models of physiological phenomena require at least two parts: the model equations to solve, and a computational domain in which to solve them.

Here, we discuss the development of a 1D flow model for arterial and venous perfusion including explicitly specified large blood vessels and vascular beds that are implicitly modelled via a structured tree approach. The model can be implemented using only physiologically meaningful parameters, or parameters that can be derived from physiological data. Further, as many blood vessels, such as the microvasculature of the coronary and pulmonary systems are embedded within moving tissues, we include the periodic external pressure to which vessels are subject to as a result of tissue motion.

As the model can be implemented with only physiologically meaningful parameters, it is reasonable that the large blood vessels are defined using physiologically realistic values, i.e. the computational domain in which we simulate blood flow resembles that of the organ system in which we wish to simulate blood flow. To do this, we require the length, radius, and hierarchical information of all blood vessels that are to be modelled explicitly. These data are difficult to find in the literature, and where they are reported, they are given in aggregate. However, imaged vascular networks are increasingly easy to obtain.

Given the need for realistic computational domains, and the increasing availability of medical images, we present the development of robust algorithms that are used to simplify large data sets that represent the vasculature of an organ system without loss of information, and pruning techniques to obtain a computational domain that conforms to given criteria required of the computational domain. In the specific example of the coronary circulation, we separate the arteries into those that lie within and outwith the myocardium. We are also able to divide the given organ into subregions that are perfused from a given large artery and compare these against the American Heart Associations division of the left ventricle as a sense-check.

Finally, we will simulate blood flow in the obtained computational domains to investigate the numerical model’s sensitivity to computational domain morphometry.

This research was supported by the UK Engineering and Physical Sciences Research Council (EP/N014642/1, EP/T017899/1) .

Human skin is a complex material to test and model as its physical and geometric properties depend on a host of parameters and conditions: thickness, location, age, ethnicity, hydration, etc. Skin tension plays a pivotal role in clinical settings, affecting scarring, wound healing and skin necrosis. Despite its importance, there is no widely accepted method for assessing in vivo skin tension. Destructive testing of skin samples only gives a partial picture, as harvesting skin dehydrates the sample and releases its residual stress, which is likely to alter its behaviour significantly. Knowledge of the in vivo tension in skin will aid in preoperative reconstructive surgery planning, e.g., by providing safe limits of skin stress or accurately estimating the skin area required for repairs. Here, we develop and validate two methods to quantify the in vivo tension in skin using acoustic measurements. We find that it is possible to determine the difference in residual stress along in-plane directions on a patient-specific basis, using a simple in vivo measurement technique. Additionally, we find that coupling elastic wave measurements with machine learning (ML) models is a viable non-invasive method to determine in vivo skin tension.

Firstly, we model the skin as an incompressible, anisotropic hyperelastic material with one family of fibres, where the principal pre-stress is aligned along the fibres. A detailed analysis of the theoretical model reveals that the in-plane stress difference is related to the surface wave speeds, via a simple formula with a known error of less than 9%. The proposed formula is universal and depends on neither a specific energy density function nor material properties. We validate the formula with finite element (FE) simulations. We replicate the in vivo stress by applying a pre-stretch and induce a wave using an instantaneous impulse. We measure the wave speeds parallel and perpendicular to the fibres at various levels of pre-stress. We find that the error between the actual stress difference and that calculated with the formula is less than 3% for the simulations, which is within the range determined by the analytic model.

Secondly, we train a ML model that uses surface wave speed measurements to predict the in vivo skin tension. We create a large dataset consisting of simulated wave propagation experiments using an FE model. We then train a Gaussian process regression model to solve the inverse problem of predicting stress and pre-stretch in skin using the wave speed measurements. The ML model demonstrated good predictive performance, highlighting the feasibility of the method.

In vivo stress is difficult to estimate in general, however, the methods proposed here will enable on-demand patient-specific measurement of the in vivo stress in skin in a non-invasive manner. They are based on easily accessible parameters and could replace existing qualitative techniques with more accurate quantitative measurements, aiding preoperative reconstructive surgery planning and ultimately improving surgical outcomes.

This is joint work with Matt Nagle, Christelle Vedel, Wenting Shu, Michel Destrade, Michael Fop and Aisling Ní Annaidh.

Age-related vision problems significantly impact the quality of life. Among these, macular holes are a leading cause of vision loss in the elderly, resulting from the formation of a hole in the macula, the central region of the retina responsible for high-acuity vision. Damage to the macula can lead to severe visual impairment. Although the precise mechanism of macular hole formation remains unclear, it is thought to be associated with mechanical forces exerted on the retina by the vitreous. The vitreous is a clear, viscoelastic gel composed of collagen fibres that occupies the space between the lens and the retina and is attached to the retina at its posterior end. With age, the vitreous undergoes structural changes and shrinks, increasing the traction forces on the retina. Simultaneously, the retinal tissue gets thinner and fragile with age. In combination, these processes compromise the integrity of the macula, cause swelling, and eventually lead to the formation of a hole in the retina. Prolonged traction forces can further result in the retina separating from the retinal pigment epithelium (RPE), a layer of epithelial cells located at the back of the retina, leading to retinal detachment. While the initiation of this pathology is qualitatively understood, a quantitative analysis of the mechanical forces contributing to macular hole formation is still lacking. It is observed that the size of the macular hole affects postoperative closure and visual acuity. Hence, a better understanding of macular hole formation and growth mechanics will help improve surgical techniques such as vitrectomy.
In this study, we introduce a poroelastic model of the retina to investigate the mechanical forces involved in macular hole formation and retinal detachment. Our model captures the age-related changes in vitreous traction forces and their impact on the retina. Using retinal OCT images as input data, we simulate the patient-specific retinal geometry with symmetry at the macula. The model incorporates a multiphasic contact framework between the two halves of the retina, ensuring continuity in fluid pressure, solute concentration, and displacement degrees of freedom. A traction-separation law-based damage criterion is employed to simulate the detachment of the retinal halves. Additionally, a multiphasic contact is defined between the retina and the RPE to simulate retinal detachment.
Our findings demonstrate that as the traction forces on the retina increase with age, the size of the macular hole also grows, consistent with clinical observations. The model further reveals how fluid and solute flows are redistributed due to macular hole formation. Moreover, it highlights the significant influence of retinal stiffness on macular hole formation and retinal detachment, suggesting a link between these conditions and other pathologies associated with increased retinal stiffness. This work provides a quantitative framework for understanding the mechanical forces driving macular hole formation and offers insights into the relationship between various retinal pathologies.

This study investigates the mechanics of avascular solid tumour growth, examining how chemical interactions in the microenvironment shape its development, with a particular focus on the delayed effects these interactions have on tissue evolution. We model the tumour's progression by using the multiplicative decomposition of the deformation gradient tensor [1], which separates the total deformation into two distinct parts, one capturing inelastic distortions caused by growth, and the other representing elastic deformations accommodating these changes.
Building on [2,3], we formulate a growth law to govern these inelastic distortions, based on a balance of generalised mechanical forces and the dissipation inequality, allowing for the inclusion of constitutive relations. This structure provides a more physically grounded framework and also permits the inclusion of mass source or sink terms from conventional phenomenological models. Furthermore, this growth law is determined a posteriori and governed by internal and external non-conventional forces. The internal force, associated with the Eshelby stress tensor, captures configurational mechanical stress arising from inhomogeneities within the tumour structure. The external force is conjugate to growth-related kinematic descriptors and models microscale biochemical interactions, such as nutrient consumption, drug treatment strategies, and other factors that may influence tumour growth and tissue behaviour over time.
A further key aspect of this study is the inclusion of time-delay effects within the growth model. In real biological systems, particularly in heterogeneous tumour microenvironments, chemical processes are not instantaneous. We account for this by introducing memory effects into the external non-conventional force through integral operators, drawing on concepts from fractional calculus, to reflect how earlier processes can continue to influence the current state of growth. Such delays become especially relevant when considering tumour response to temporally varying environmental inputs, including nutrient supply or therapeutic interventions. These time-dependent effects improve the model's ability to reflect tumour progression beyond time-local formulations and offer a mathematically consistent way to account for non-instantaneous biochemical effects.
Our findings show how time-delay effects, combined with the growth law, affect key tumour descriptors over time. This approach highlights the connection between delayed chemical influences and a mechanically grounded growth law, offering a new perspective on tumour mechanics and contributing to a deeper understanding of the complexity of the tumour microenvironment.
References
[1] Micunovic, M. (2009). Thermomechanics of Viscoplasticity. Springer. https://doi.org/10.1007/978-0-387-89490-4
[2] Grillo, A., & Di Stefano, S. (2023). Mathematics and Mechanics of Complex Systems, 11(1), 57–86. https://doi.org/10.2140/memocs.2023.11.57
[3] Almudarra, M., & Ramirez-Torres, A. (2025). Mathematics and Mechanics of Solids, 30(2), 501–526. https://doi.org/10.1177/10812865241230269

We study numerically the conformations of self-avoiding active ring polymers. They are modeled as closed bead-spring chains, driven by tangential forces, whose dynamics is implemented via the Brownian multiparticle collision scheme.
The mean-square intrachain distances, the mean-square gyration radius, and the bond correlation functions are calculated, analyzed using scaling assumptions, and comprared to the same quantities of the corresponding passive case.
We show that all these quantities can be described asymptotically by the same scaling forms of passive ring polymers but with a new active scaling exponent.
At high activity it is found that an effective persistence length characterizes the conformations of flexible active rings.

Investigating intracardiac flow dynamics is crucial for diagnosing and treating congenital heart defects, as flow patterns can indicate disease progression and inform the timing of therapeutic interventions. While 4D flow MRI enables non-invasive quantification and visualization of the cardiovascular flow field, its clinical applicability is constrained by limited spatial and temporal resolution, which restricts its capability to resolve fine-scale flow features and derived hemodynamic parameters such as vorticity, wall shear stress, and energy loss. These limitations are especially important in pediatric cases, where small anatomical scales and rapid heart rates requires higher resolution. Computational tools can fill this gap and while significant progress has been made in developing advanced multi-physics computational models that account for the complex interactions of elasto-mechanics, electrophysiology and hemodynamics in the heart, their direct application to patient specific cases in congenital heart disease remains problematic. This is due to their dependance the constitutive relations used for example to model fluid-solid interactions that rely heavily on experimentally derived material properties, which may not accurately represent patient-specific or pathological conditions.

In this work we propose a cost-effective alternative that reduces model complexity by replacing multiphysics components with prescribed kinematics derived from patient-specific imaging data. This approach preserves flow-related information and resolution while drastically reducing computational costs, making it more suitable for clinical applications. It also facilitates model adaptation to both healthy and pathological cases, offering advantages in pediatric cardiology, where rapid, patient-specific analysis is essential. Despite these benefits, reconstructing patient-specific geometries—particularly for thin, dynamic structures like the tricuspid valve remains a significant challenge. The delicate leaflets are often poorly resolved in conventional imaging modalities due to temporal and spatial resolution constraints. Although previous studies have proposed structural models of the tricuspid valve with varying complexity, most rely on experimental material properties that are difficult to individualize, limiting their applicability in patient-specific scenarios. Here we present a novel, computationally efficient model derived entirely from 4D-MRI velocity data. Our approach is based on the observation that the gradient of the velocity magnitude is highest near vascular boundaries, including the tricuspid valve leaflets, due to impermeability. Segmentation and registration of the right ventricle (RV) and right atrium (RA) were performed using 3D Slicer, an open-source platform for medical image processing. The final kinematic reconstruction of the RV, including the valve geometry, was prescribed using the Large Deformation Diffeomorphic Metric Mapping (LDDMM) framework.

We conducted direct numerical simulations (DNS) using an in-house solver based on the Immersed Boundary Method (IBM) to solve the Navier–Stokes equations for incompressible flows. The 4D-MRI velocity field served as a benchmark, and DNS results were compared both qualitatively (velocity field during diastolic filling) and quantitatively. Our findings demonstrated strong agreement between DNS and 4D-MRI, particularly in capturing diastolic RV filling patterns in patient-specific cases. This validates our model’s ability to replicate physiological flow without relying on expensive multi-physics simulations, offering a scalable tool for clinical hemodynamic assessment.

To tackle some of the current challenges in robotics, such as adaptability in complex and unpredictable environments, as well as miniaturization constrains, scientists have engineered bio-bots, by integrating living cells with synthetic components. Beyond advancing robotics, these biohybrid platforms offer valuable insights into the fundamental biophysics and biology of natural systems, with potential applications in human health and medicine. Despite their promising capabilities, significant challenges remain, including ethical implications, technical roadblocks in control strategies and the development of predictive computational modelling tools to accurately design, simulate and optimize the biohybrid prototypes.

Taking inspiration from the biophysics of the human heart, we have conceived a new class of biohybrid swimmers. The envisioned system consists of a hollow cylinder made of cardiac muscle cells (cardiomyocytes), integrated with an artificial activation device. As in the human heart, the sinoatrial node triggers the synchronized contraction of muscle fibres to enable rhythmic blood pumping, the propulsion mechanism of our prototype is driven by an electric current generated by a pacemaker situated on its leading edge. This impulse trigger periodic muscular contractions, resulting in the propagation of peristaltic elastic waves along the soft body of the swimmer, providing the thrust for the forward locomotion, in the low to intermediate Reynolds number regime.

More specifically, the motion of the swimmer results from the interconnected dynamics of the electrophysiology system, responsible for the action potential propagation that triggers the active muscular tension, the internal passive forces of the biological tissue, and the hydrodynamic loads due to the interaction with the surrounding fluid. By simultaneously accounting for the electrophysiology, elastomechanics, and hydrodynamics, as well as their interaction, our state-of-the-art computational tool allows us to optimize and predict the swimming dynamics through three-dimensional numerical simulations.

In principle, the swimming characteristics, optimal configuration, excitation frequency and physical specifications of the system are highly dependent on the specific condition being simulated. However, our findings demonstrate that the system exhibits unexpectedly robust features. Each configuration analysed is characterized by an optimal excitation frequency, maximizing both swimming speed and initial acceleration. Notably, the optimal actuation frequency is Reynolds-independent, indicating that viscosity does not significantly impact the propulsion mechanism and is therefore not a key parameter in the design of such structures.

To elucidate the behaviour of the system, we propose a physics-based analytical model capable of accurately predicting the swimming dynamics in the optimal scenario. Interestingly, our findings show that peak performance is achieved when a single peristaltic wave propagates along the swimmer body, whereas existing models in the Stokes regime predict that increasing the number of pulses enhances swimming speed. In particular, once the contraction reaches the distal end, the peristaltic cycle immediately resumes, providing uninterrupted propulsion. Finally, we capture the fundamental aspects of the system’s locomotion strategy by deriving a universal scaling that links the swimming speed to body kinematics.

In conclusion, the proposed proof-of concept provides useful design guidelines for the development of future biohybrid devices, which can lead to new insights in robotics, bioengineering and medicine.

Nowadays, numerical simulations are valuable tools for cardiovascular research. They allow for testing new devices and surgical interventions and evaluating quantities difficult to measure in vivo or in vitro. Nevertheless, the reliability of their results is strongly linked to the assumptions and uncertainties in the computational model. The selection of an appropriate constitutive relation for blood viscosity represents a crucial concern for computational models of haematic flows.
Blood is a biphasic fluid consisting of a solid phase, which includes red blood cells (RBCs), white blood cells (WBCs) and platelets, and a liquid phase, i.e. plasma. Viscoelastic, shear-thinning, and thixotropic properties of blood are mainly attributed to the dynamics and interactions of RBCs. Therefore, non-Newtonian rheology should be taken into account when modelling blood flows. However, when strains in the flow are higher than a given threshold, considering blood as a Newtonian fluid is a reasonable hypothesis. For this reason, several previous studies adopted this assumption, especially for flows in large vessels. Nonetheless, it may not be adequate under pathological conditions and at instants of the cardiac cycle when recirculation regions are present. Previous works compared Newtonian and non-Newtonian rheological models for flows through heart valves, arteries, and single cardiac chambers, often introducing some simplifications. In the present study, we perform numerical simulations of the fluid-structure-electrophysiology interaction (FSEI) in the left heart with an in-house developed, GPU-accelerated code. We assume the Newtonian and Carreau constitutive law for blood viscosity. We assess the influence of the rheological model on haemodynamic quantities of interest for clinical practice. Specifically, non-Newtonian rheology does not substantially affect physical quantities depending on the volumetric stress tensor, such as the pressure in the chambers and aorta or the orifice area of the valves. On the other hand, we observe considerable differences in the wall shear stress and local haemolysis between the Newtonian and non-Newtonian models. Setting the constant viscosity in the Newtonian model equal to the spatio-temporal average of the value estimated by the Carreau model does not mitigate the discrepancies. For this reason, non-Newtonian rheological models are recommended if interested in red blood damage, tissue remodelling and atherogenesis.
Future work aims to include more complex non-Newtonian properties in the rheological model, such as thixotropicity and viscoelasticity. Moreover, since previous studies highlighted that non-Newtonian effects are more evident when diseases are present, the analysis will be repeated for a left heart under pathological conditions. Additionally, the effect of haematocrit will also be investigated with the goal of improving the accuracy of patient-specific simulations.

The contraction of the heart during a heartbeat is the consequence of the synchronized reaction of the myocardium to the propagation of a non-linear electrophysiology wave. Such scenario, however, can be altered in pathological conditions, such as when ventricular tachycardia or ventricular fibrillation manifest. These phenomena are related to the non physiological tissue activation under the propagation of non-planar electrical waves, namely spiral and scroll waves, which are often triggered by the presence of defects introduced by myocardial infarction. In particular, the scar region is known to act as a support for the generation of scroll waves owing to the high conductivity difference between healthy and pathological tissue. While the dynamics and formation of arrhytmogenic patterns is widely studied in literature [ 1, 2], their consequences on the cardiac hemodynamics are rarely taken into account. In this work, we aim at investigating what are the effects of ventricular tachycardia on the more relevant bio-markers, in particular, determine what pathological electrophysiology patterns have the most significant impact on the physiological blood flow dynamics in the left heart and try to understand the optimal treatment. To this aim, the tissue activation data of a patient affected by an arrhythmogenic ventricular scar mapped through the CARTO^® 3 system, has been used in order to fine tune the heterogeneities of the substrate for the electrophysiology model in order to trigger reentries in the myocardium. The electrophysiology is then solved trough a monodomain model coupled with suitable cellular models encompassing the hetoregeneity of the myocardium owing to the presence of
healthy tissue, peri-infarct region and scar tissue. The corresponding hemodynamics is then solved by integrating the Navier-Stokes equations, which are discretized using a staggered finite differences complemented with immersed boundary techinques [3]. The resulting multi-physics system can be then exploited as a predictive tool to reproduce ventricular arrhythmias and try to understand what is the optimal intervention in order to restore the physiological cardiac function.

¹ Salvador M. et al., Computers in Biology and Medicine, 142, 2022.

² Ramírez W. A. et al., Scientific Reports, 10, 2020.

³ Viola F. et al., Scientific Reports, 13, 2023.

Stress experienced by a mother during pregnancy can profoundly influence the neurodevelopment of the fetus, potentially increasing susceptibility to autism spectrum disorders. We investigate these effects by extending a minimal mathematical model of the hypothalamic–pituitary–adrenal (HPA) axis to include mirror neuron activity in the premotor cortex. Our coupled system of differential equations captures the short-term (acute) versus long-term (chronic) impacts of maternal stress on embryonic development. We show that acute stress causes transient elevations in cortisol and mild alterations in mirror neuron activity, whereas chronic stress yields persistent physiological changes—elevated adrenal gland mass, higher baseline cortisol levels, and diminished mirror neuron activity that can remain even after the stressor is removed. Through phase-portrait and fixed-point analysis, we identify a key parameter, D, which governs the dynamics of mirror neuron activity. When D is small, three fixed points arise, including two stable equilibria—one corresponding to a “healthy” (euthymic) state and the other associated with reduced mirror neuron activity (an “autistic” state). Under large D, the system supports only the euthymic fixed point, indicative of a non-susceptible population. Finally, we demonstrate that while acute reductions in stress have negligible impact on the long-term dynamics, chronic reduction of stress can restore cortisol levels and mirror neuron activity to the euthymic range. These findings suggest that chronic maternal stress can be a critical determinant in the onset of autism-like neurodevelopmental trajectories for susceptible individuals and underscore the importance of sustained stress management interventions during pregnancy.

Integrating biological control with species conservation offers a sustainable solution for pest management while promoting biodiversity. This study examines a community-led conservation effort in Ayeyarwady Delta, Myanmar, where the invasive Golden Apple Snail (GAS) has severely damaged rice crops. The chemical controls and manual removal have been costly and environmentally harmful. In response, the Kyonekapyin-Tapseik Community Conservation Group (KTCG) in the delta area has focused on protecting the Asian Openbill stork (Anastomus oscitans), a natural predator of GAS, as part of a nature-based solution. We asked structured questionnaires to the farmers (n= 92) in eight villages. The results show a significant decline in GAS populations and a notable recovery in rice yields since 2018 when the stork protection has been started. Farmers observed that the storks’ foraging habits effectively reduced snail infestations. Before GAS infestations, average rice yields metric tonne per hectare were 0.73 in monsoon and 1.23 in summer, dropping significantly during infestations 0.15 and 0.85 in monsoon and summer, respectively, and restoring back to 0.55 and 1.21 in monsoon and in summer by protecting Asian Openbills which suppress GAS. The farmer attitudes survey (n=92) result shows that conserving biodiversity is important by strongly agree (n=89), agree (n=2), neutral (n=0) and disagree (n=1). None of farmers response for strongly disagree. This study highlights the dual benefits of conservation-driven pest control: improved agricultural productivity and species protection. By safeguarding the Asian Openbill stork, farmers not only reduced crop losses but also helped conserve a threatened species.

The rapid and continuous evolution of digital markets has significantly amplified the demand for intelligent, data-driven decision-making tools that not only enhance customer engagement but also improve operational efficiency and long-term business performance. In the highly dynamic landscape of digital commerce, businesses are increasingly faced with the challenge of managing large volumes of unstructured consumer data while making timely, insightful decisions. The complexity of customer interactions, particularly within e-commerce platforms, requires innovative solutions that can extract meaning, predict behavior, and support customer retention. In this context, the integration of Fourth Industrial Revolution (4IR) technologies such as artificial intelligence (AI), machine learning (ML), and intelligent automation has become indispensable for achieving competitiveness and driving strategic innovation.

This study presents a novel application of 4IR technologies, with a focus on machine learning and sentiment analysis, aimed at transforming digital business practices. Using Amazon e-commerce review data as a case study, we designed an intelligent framework that leverages advanced feature engineering, Synthetic Minority Oversampling Technique (SMOTE) for class balancing, and the AdaBoost ensemble learning algorithm. This integrated approach was developed to address key digital business objectives namely, improved sentiment classification and accurate prediction of customer churn.

We implemented and evaluated multiple classifiers including Random Forest, Logistic Regression, Extra Trees, and XGBoost, training them initially on baseline data and then enhancing them using SMOTE, AdaBoost, and their combined application. To comprehensively assess the models' performance, we employed several key metrics: accuracy, precision, recall, and Matthew’s Correlation Coefficient (MCC). Results demonstrate that AdaBoost significantly improves classification accuracy, with marked precision gains observed in Extra Trees and XGBoost. SMOTE was particularly effective in improving recall rates, especially when used with XGBoost. Notably, the combined use of SMOTE and AdaBoost resulted in enhanced MCC scores for both Random Forest and XGBoost models, indicating a more balanced and reliable predictive performance.

These findings underscore the transformative potential of integrating 4IR techniques into digital business systems. Our proposed framework proves valuable not only in enhancing the precision of sentiment analysis but also in enabling proactive management of customer churn two key challenges in digital business today. By providing actionable insights, the framework empowers organizations to make more informed, agile, and customer-centric decisions. This contributes to improved customer loyalty, reduced attrition, and more resilient digital business strategies.

Future research will explore the real-time deployment of the model in dynamic e-commerce environments and its extension to multilingual and multimodal datasets to ensure broader applicability and effectiveness across diverse global markets. This research ultimately contributes to the growing body of knowledge on digital transformation within the Industry 4.0 paradigm.

Medium-size and Large-bodied Mammals of Marshall Wetlands, a Proposed Protected Area in Liberia.

Abstract. Liberia comprises 43% of the upper Guinean tropical forest of West Africa, a global hotspot with high levels of biodiversity and endemism succumbing to accelerated rates of deforestation. Due in part to the country’s past 14 years of civil war, knowledge about much of Liberia’s remaining local biodiversity is limited. To help fill this gap, we surveyed medium-sized and large-bodied mammals in a proposed protected area, the Marshal Wetlands, in Margibi County, Liberia. We used 21 camera traps (Bushnell HD model 119739), installed at a minimum distance of 250 meters from one another, and at a height of 60 cm from the ground floor. Camera traps were set to take a sequence of 3 pictures every 2 minutes, using the highest quality setting both for the photos and sensor, to cover the entire trail. Trails included the three main habitats present in the area, primary and secondary forests, mangroves, and nearby human settlements. In an effort of 33,120 hours, and track surveys along trails, between July 2022 and May 2023, we confirmed the presence of 13 medium-sized and large mammal species. Of these, four species have some degree of threat in the IUCN Red List. Despite its proximity to the capital city of Monrovia, the Marshall Wetlands comprise significant biodiversity, highlighting its conservation value and reinforcing the value of improved protected status.

Key words. Biodiversity, camera-trapping, mangrove, species list, tropical forest, West Africa

Gweh A, Bailey A, Palmeirim AF, Poultolnor J, Toe H, Crowley W, Machalaba C, Desmond J, Desmond J,

Prist PR (2025) Medium-sized and large-bodied mammals of Marshall Wetlands, a proposed protected area

in Liberia. Check List 21 (1): 1–11. https://doi.org/10.15560/21.1.1

Acute Respiratory Distress Syndrome (ARDS) is a critical condition. For patients suffering from this pathology, mechanical ventilation (MV) is needed to ensure sufficient ventilation and oxygenation. Intensivists have several therapeutic tools at their disposal to design an adapted MV. However, they still face several issues in their decision-making when designing it:

• High variability between patients
• Limits in the understanding of the underlying mechanisms

That is why MV design is still mostly experience-based. In order to improve the understanding of ARDS lung biomechanics and physiology, a coupled physio-mechanical computational framework was conceived and implemented.

The model is a lumped-parameter model which is conceptually simple and has a limited number of degrees of freedom in order to keep the computational burden low. The discretization of the lung is made into 2N compartments called Respiratory Units (R.U.), with N being the number of generations of the trachea-bronchial tree considered. The geometry and physical properties of each R.U. are based on the patient's CT-scan and lung segmentation. Throughout a simulation, two submodels are run: the mechanical and the physiological ones. In the mechanical model, each R.U. is able to inflate and deflate according to the input MV and a non-linear compliance. The R.U. can also collapse and reopen to embody atelectasis. In the physiological model, gas exchanges (O2 and CO2) consequent to the inflation computed in the mechanical model are calculated for each R.U. This model allows the simulation of the patient's oxygenation, including phenomena such as hypoxic vasoconstriction. Notably, the coupling between the two submodels allows for the simulation of the ventilation-perfusion mismatch, which is data of high clinical significance.

First results showed that the model behaves as expected qualitatively. Clinically observed phenomena were reproduced. For instance, from a biomechanical point of view, recruitment in dorsal areas and lung homogenization when positioning the patient prone were reproduced. Concerning the physiological model, the discrepancy between tidal volume increase and oxygenation improvement was simulated and explained.

These results are promising and demonstrate the ability of the model to capture part of the ARDS lung biomechanics and physiology. However, the model's validity is currently not validated against patient data, and in its current form, it provides qualitative results only. Future work will aim first to identify the most influential parameters through a global sensitivity analysis. Then, an inverse method framework will be developed to tune this reduced set of parameters based on patient data to make quantitative predictions

Pulmonary capillary perfusion and gas exchange are critical physiological processes that occur at the alveolar level and are fundamental to sustaining life. Current computational simulations of these phenomena often rely on simplified, low-dimensional mathematical models that do not directly account for the alveolar morphology [1]. These models typically involve simplified representations of the complex chemical reactions between O2​-CO2​ and hemoglobin. While such models offer general insights, they frequently fail to capture the intricate nature of chemical reactions within pulmonary capillary blood flow in realistic geometries. Additionally, they overlook the significant impact of microstructural morphology on overall pulmonary function. In this work, we introduce a coupled continuum model to simulate pulmonary perfusion and gas exchange in alveolar geometries [2]. We further include complex gas and hemoglobin dynamics within realistic geometries of alveolar tissue [3]. To achieve this, we derive a set of governing equations that include a two-way Hill-like relationship to describe the interaction between gas partial pressures and hemoglobin saturations. We employ a non-linear finite-element approach to solve the resulting boundary-value problem numerically. This approach enables us to simulate and validate key physiological fields such as blood velocity, oxygen and carbon dioxide partial pressure, and hemoglobin saturation in idealized and complex geometries. We also conduct sensitivity studies to examine the influence of blood speed and acidity variations on these key physiological fields. Notably, we extend our simulations to anatomical alveolar domains reconstructed from 3D computed-tomography images of murine lungs. Our simulations reveal that morphological variations within these domains lead to significant changes in O2​ and CO2​diffusing capacity, which can be related to pulmonary chronic diseases.

Acknowledgements: This work was supported by the Agencia Nacional de Investigación y Desarrollo, Chile, through grant FONDECYT 1220465.

[1] Clark, A. R., Burrowes, K. S., & Tawhai, M. H. (2021). Integrative computational models of lung structure-function interactions. Comprehensive Physiology, 11(2), 1501–1530. https://doi.org/10.1002/cphy.c200011

[2] Zurita, P., & Hurtado, D. E. (2022). Computational modeling of capillary perfusion and gas exchange in alveolar tissue. Computer Methods in Applied Mechanics and Engineering, 399, 115418. https://doi.org/10.1016/j.cma.2022.115418

[3] Herrera, B., & Hurtado, D. E. (2025). Modeling pulmonary perfusion and gas exchange in alveolar microstructures. Computer Methods in Applied Mechanics and Engineering, 433, 117499. https://doi.org/10.1016/j.cma.2024.117499

Introduction: Computational lung modeling has become an essential tool for virtually analyzing respiratory mechanics and disease progression in individual patients, providing detailed insights into key variables such as deformation, stress, and pressure distributions [1]. Despite its advantages, the high computational cost of current organ simulations, often requiring several hours, restricts their use in clinical settings. To overcome this challenge, machine learning (ML) has shown promise in accelerating predictions through surrogate models. However, most existing approaches are limited to the study of structured grids, fixed domains, or parameterized geometries. Graph Neural Networks (GNNs) have recently gained attention for their inherent ability to adapt to intricate and irregular geometries. In this study, we propose a GNN-based framework to develop efficient spatiotemporal surrogates of finite element lung models under mechanical ventilation. Our approach is designed to handle diverse and complex lung geometries, while remaining generalizable to patient-specific boundary conditions.

Materials and methods: Using CT scans, we created a dataset consisting of N different 3D lung meshes from different patients {M₁,M₂,…,M_N} and conducted finite element simulations of lung behavior under mechanical ventilation [2]. The simulation outputs included the spatiotemporal displacement and alveolar pressure fields over a ventilation cycle of duration T. To accelerate these model predictions, we developed and trained a Graph Neural Network (GNN) architecture named LungGraphNet (LGN). This model adapts each patient-specific lung mesh into a computational graph structure G=(V,E), where V represents the nodes and E the corresponding connecting edges. The physical boundary conditions essential to describe the problem are embedded as nodes and edges features into this graph representation. Once trained, LGN serves as a surrogate model capable of performing inference on a new and unseen lung geometry G_i​, accurately predicting the spatiotemporal response and reconstructing the evolution of respiratory signals under mechanical ventilation.

Results: The LGN displayed high accuracy in predicting spatiotemporal displacement and alveolar pressure fields, achieving a relative error of less than 5% across all patient cases in the test set, compared to the ground-truth simulation values. During inference, the surrogate model generated predictions for arbitrary lung geometries in less than 2 seconds, which is approximately 4800 times faster than the finite element simulations and without requiring retraining.

Discussion and Conclusions: The proposed framework offers an efficient solution for patient-specific lung modeling. The significant speed-up achieved by our surrogate model makes it a promising tool for the potential use of virtual lung analysis in both clinical practice and research.

References:

1. Neelakantan S et al. (2022), JR Soc. Interface. 19(191), 20220062

2. Avilés N, Hurtado D (2022), Frontiers in Physiology. 13, 984286.

Acknowledgements: The authors would like to thank the National Agency for Research and Development (ANID), graduate fellowship ANID BECAS/DOCTORADO NACIONAL #21220063, and Pontificia Universidad Católica de Chile, for providing financial support to this project.

The lung is a complex system that serves as an interface for the exchange of gases between the ambient air and the bloodstream. Its main function is to oxygenate the blood and remove carbon dioxide through its tree-like structure. Once oxygen enters the bloodstream, hemoglobin becomes its principal carrier. However, the affinity of hemoglobin for oxygen is influenced by the concentration of carbon dioxide in the blood, leading to a competition between the two gases to bind with hemoglobin. This phenomenon is known as the Haldane and Bohr effects.

The study of gas transport mechanisms and respiratory gases exchange with blood is a fundamental step to better understand how this complex system works. For this, a gas transport model has been developed. It is driven by the applied pressure on the lung and is composed of advection-diffusion-reaction PDEs for each gas species in one dimension. This model, contrary to its 0D counterpart [1], naturally accounts for the mixing of gases along the bronchial tree and introduces a time delay as the gases have to be transported before reaching the alveoli. Furthermore, it also takes into account the strong interaction between oxygen and carbon dioxide in the blood thanks to a coupled ODE system that represents the nonlinear coupling in the diffusion of the gases in the blood.

This model ultimately allows for the computation of the amount of oxygen and carbon dioxide exchanged with the blood over time along with their arterial partial pressures. The data predicted by the model are fully consistent with the physiological responses expected in the healthy case. Additionally, we explore how both the 1D model and its 0D counterpart respond to different breathing patterns by examining two types of applied pleural pressure—piecewise constant and piecewise exponential—across various breathing periods, inspiratory ratios, and pressure amplitudes.

Moreover, this model is used to consider optimal control scenarios by minimizing a cost function in order to find optimal breathing scenarios. This involves exploring which cost functions the observed stereotypical breathing scenario may optimize. The underlying assumption is that optimal breathing should be a combination of several criteria: low effort, minimal lung distension, while maintaining the carbon dioxide arterial partial pressure at a given level. By investigating these cost functions, the model can provide insights into the most efficient breathing patterns that balance these criteria, further enhancing our understanding of respiratory physiology and paving the way for more effective breathing therapies.

[1] L. Boudin, C. Grandmont, B. Grec, and S. Martin. A coupled model for the dynamics of gas exchanges in the human lung with Haldane and Bohr’s effects. Journal of Theoretical Biology, 573:111590, 2023.

Pelvic organ prolapse (POP) is a condition that significantly impacts the quality of life for many women and is characterised by the descent of pelvic organs. Initially, synthetic polypropylene meshes were introduced for POP repair due to their high success rates in correcting abdominal hernias. However, the Food and Drug Administration (FDA) later banned the commercialisation of these meshes, highlighting the urgent need for alternative approaches [1]. This study aims to develop biodegradable meshes as an alternative solution for POP repair. Biodegradable meshes offer promising biocompatibility and biomechanical properties, making them a potential solution to the limitations of synthetic meshes.

Computational models were utilised to design meshes with variations in pore geometry, pore size, filament thickness, and filament placement in specific regions. To validate the simulation results, meshes were 3D printed, and uniaxial tensile tests were conducted using the sow's vaginal tissue. The results were compared with simulations to identify meshes exhibiting behaviour similar to vaginal tissue. Additionally, outcomes were compared with the properties of the uterosacral ligament and a commercially available mesh [2].

The mesh featuring a smaller pore diameter (1.50 mm), strategically placed filaments, and variable filament thickness most closely replicated the behaviour of vaginal tissue. However, none of the tested meshes exhibited behaviour comparable to the uterosacral ligament. The findings also suggest that the commercially available mesh may not be the optimal treatment option, as it does not accurately represent the behaviour of either vaginal tissue or the uterosacral ligament [2].

With these findings, a new concept emerged, developing meshes using different materials with varying degradation periods, such as polylactic acid (PLA) and polycaprolactone (PCL). Given the distinct behaviours of these materials, it would be interesting to develop meshes using both materials with the previously designed geometries. By varying the proportions of PLA and PCL, it will be possible to create meshes that closely mimic the behaviour of vaginal tissue and ligaments. Preliminary studies indicate a decrease in Young's modulus, suggesting promising potential for mesh development. In this initial test, meshes with square pores were developed using filaments composed of 96% PLA and 4% medical-grade PCL to assess the material behavior, yielding a Young's modulus of approximately 40 MPa. Investigating the interaction between these materials and their impact on overall biomechanical performance may lead to innovative solutions for more effective, patient-specific POP repairs.

REFERENCES

[1] FDA, “FDA takes action to protect women’s health, orders manufacturers of surgical mesh intended for transvaginal repair of pelvic organ prolapse to stop selling all devices,” https://www.fda.gov/news-events/press-announcements/fda-takes-action-protect-womens-health-orders-manufacturers-surgical-mesh-intended-transvaginal.

[2] M. F. R. R. Vaz, M. E. Silva, M. Parente, S. Brandão, and A. A. Fernandes, “3D printing and development of computational models of biodegradable meshes for pelvic organ prolapse,” Engineering Computations (Swansea, Wales), vol. 41, no. 6, pp. 1399–1423, Aug. 2024, doi: 10.1108/EC-12-2023-0967.

Understanding the mechanical properties of the spinal dura mater is essential for accurate computational modeling of the spinal cord complex. Despite its importance, characterizing this membrane poses unique challenges due to its high anisotropy, thin geometry, and sensitivity to various experimental conditions. This study present a comprehensive evaluation of how measurement technique selection and test protocols - specifically, digital image correlation (DIC) and paint application for speckle generation, crosshead displacement tracking, and preconditioning - impact the resulting mechanical parameters of human spinal dura mater.

In this experiment, spinal dura mater was collected from human donors, from which standardized samples were cut in both longitudinal and transverse orientations. Some of the specimens were coated with a high-contrast paint to enable DIC, while the remaining ones were measured conventionally, based on the crosshead displacement of a universal testing machine. In addition, selected specimens underwent preliminary cyclic loading (preconditioning). The stress in the samples was expressed in terms of I Piola-Kirchhoff stress (PK1) relative to the stretch ratio, and the results were statistically analyzed with consideration of potential demographic and anthropometric factors. Poisson’s ratio was determined from the ratio of the Hencky (logarithmic) strain measured in the transverse direction to that in the longitudinal direction.

Comparisons between crosshead displacement and DIC revealed that the choice of measurement approach significantly influences the extracted mechanical parameters. For longitudinal specimens measured by DIC, elastic moduli were up to 70% higher than values derived from crosshead data, while ultimate strain was up to 35% lower. Further, paint application itself altered tissue behavior, particularly affecting stiffness. In contrast, preconditioning did not emerge as a strong factor in modifying the stress-stretch results. After applying correction factors for these effects, we obtained averaged values for longitudinal extension of approximately 282 MPa (elastic modulus), 25 MPa (failure stress), and 1.10 (failure stretch). In the transverse orientation, these parameters were around 15.6 MPa, 1.56 MPa, and 1.12, respectively.
Notably, Poisson’s ratios ranged from about 2.6 to 4.3, underscoring the tissue’s anisotropic and potentially volume-altering behavior under load.

The findings highlight the necessity of carefully standardizing experimental protocols for testing anisotropic tissues like spinal dura mater. Methodological choices - particularly regarding strain tracking (DIC vs. crosshead), sample preparation, and paint application - can significantly influence stiffness and ultimate strain measurements. These results provide input parameters and correction guidelines for computational models, aiding in the development of more reliable simulations of spinal biomechanics.

Introduction

Dynamic elastography is a fundamental technique to study the local mechanical property of the tissues, such as cornea. It is based on in-vivo tracking of shear waves propagation as a result of a transient stimulation. In quasi-incompressible materials, as the cornea, the shear waves are 150 times slower than the compressional waves. The quasi-incompressible behavior and the double-scale of the phenomena make the FE approximation difficult. We used an efficient scheme to obtain a reliable modelling of transient elastography measurements in incompressible pre-loaded tissues, and we applied to the cornea to reproduce a typical dynamic elastography experiment, on cornea with and without defects.

Method

Wave propagation in cornea can be treated as a linear perturbation of an already loaded material, under the assumption that the wave amplitude is small. The specificity of the biological tissues is the complexity of their mechanical response of tissue, which is almost incompressible, hyperelastic and anisotropic.

We performed a linearization of the elastic problem around the prestressed state. The linearized rigidity has then 2 components: a material one, due to the non-linearity of the constitutive law, and a geometric one, associated with the prestress. For the material constitutive law, we used a microsphere model including the contributions of the isotropic matrix, the collagen lamellae and the quasi-incompressibility of the tissue (Giraudet et al. 2022).

The simulations were done using a mesh reproducing the geometry of the cornea (Pandolfi et al. 2008). The static non-linear problem is solved through a classical iterative method. The wave propagation simulation presents a challenge due to the quasi-incompressible behavior: the time-step is controlled by the velocity of the fastest wave, the compression one, while we are interested in the slow, shear, waves. To overcome this difficulty, we use a fully explicit numerical method (Merlini et al. 2025), based on high-order spectral elements, mass-lumping together with Gauss-Lobatto quadrature rule and an inf-sup stable mixed formulation, and a relaxation of the CFL condition through the use of Chebyshev polynomials.

Results

We observe that the fibers have a limited effect on the main wave velocity, despite the increase of stiffness in the cornea. This may be due to our approximation with they are mainly in-plane, which is consistent with cornea observations, even if some out-of-plane orientations are reported (Petsche et al. 2013). However, they impact significantly a faster wave, with a much lower amplitude.

The IOP and the associated prestress has a significant impact on the velocity of the main wave: the wave is faster on a cornea under pressure than in a cornea in which the IOP has been suppressed.

The introduction of a mechanical defect leads to a significant alteration of the shape of the wave front, indicating that this method can be used for the early detections of pathologies such as keratoconus.

References

- C. Giraudet et al., JMBBM, 129:105121, 2022.

- A. Pandolfi et al., J Biomech Eng, 130: 061006, 2008.

- G. Merlini et al., Waves, submitted 2024.

- S. Petsche et al., BMMB, 12:1101–1113, 2013

Hernia is a physiological condition that significantly impacts patients’ quality of life, where there is an organ prolapse through the wall of the cavity that is normally contained, due to a weakness or opening, mainly of the abdominal wall. Surgical treatment for hernias often involves the use of specialized meshes to support the abdominal wall. While this method is highly effective, it frequently leads to complications such as pain, infections, inflammation, adhesions, and even the need for revision surgeries. According to the Food and Drug Administration (FDA), hernia recurrence rates can reach up to 11%, surgical site infections occur in up to 21% of cases, and chronic pain incidence ranges from 0.3% to 68%. These statistics highlight the urgent need to improve mesh technologies to minimize such complications.
In this study, a preliminary innovative melt-electrowritten mesh with antistatic properties from hernia repair is presented. Melt Electrowriting is a promising technique, which allows a precise deposition of thin fibers, mimicking the fibrillar component of the native extracellular matrix. The study investigated the effect of incorporating the antistatic agent Hostastat ® FA 38 (HT) at concentrations of 0.03, 0.06, and 0.1 wt% on the fiber diameter and mechanical properties of Polycaprolactone (PCL) meshes. The addition of HT reduced fiber diameter by 14–17%, 39–45%, and 65–66%, depending on the mesh geometry (square or sinusoidal, both with a 1.5 mm pore size). The reduced fiber diameter correlated with increased tensile strength and Young’s Modulus, as verified through uniaxial tensile testing. Comparisons between PCL/HT and pure PCL meshes with similar diameters and geometries confirmed the enhanced mechanical properties of PCL/HT meshes. Cytotoxicity tests, using the resazurin assay, indicated no cytotoxic effects at any HT concentration. Both sterilization methods (Ethanol + UV and UV only) showed no significant differences in results. These findings demonstrate that incorporating HT into PCL meshes produces thinner, more stable filaments with improved mechanical performance and no cytotoxicity, making them a promising material for applications such as hernia repair. Future studies on polymer-additive interactions could further optimize their mechanical and biological properties.

Keywords: Hernia Repair; Biodegradable Mesh Implants; Melt Electrowriting; Antistatic Agent; Cytotoxicity

Biodegradable magnesium (Mg) implants offer an attractive alternative to permanent metallic devices by gradually dissolving as natural bone regenerates, thereby reducing long‐term complications. However, predicting the degradation behavior of Mg and its interaction with surrounding tissues remains a significant challenge. In our work, we present a novel multiscale computational framework that couples a nonlocal peridynamic corrosion model with a bone healing and remodeling model. The peridynamic model accurately captures the time-dependent degradation of Mg by simulating its progressive dissolution and macroscopic volume loss, while dynamically linking the release of Mg ions to local biological responses. This ion release directly influences osteoblast differentiation and subsequent mineralization of the bone matrix, providing new insights into the connection between implant degradation and tissue regeneration. While our primary focus is on quantifying the biocorrosion of Mg and its immediate impact on bone healing, our framework also lays the groundwork for future integration of soft tissue biomechanics to further enhance implant-tissue integration. This coupled model serves as a robust tool for designing next-generation biodegradable implants and optimizing their mechanical performance and biocompatibility. Our findings promise to advance implant design strategies by linking degradation kinetics with tissue growth, thereby facilitating a more predictive and personalized approach to clinical treatment and opening avenues for interdisciplinary research in biomechanics and regenerative medicine.

The stomach plays a central role in digestion by accommodating ingested food, mechanically mixing its contents, and chemically breaking down nutrients. These functions are tightly orchestrated by gastric peristalsis—a complex process governed by an intricate electromechanical system that ensures the effective propulsion of food into the intestines. A detailed understanding of the mechanics and regulation of gastric motility is therefore crucial for diagnosing and treating prevalent gastrointestinal disorders such as gastroparesis, gastroesophageal reflux disease (GERD), and functional dyspepsia, all of which can significantly impair quality of life.

Despite its clinical relevance, gastric electromechanics remain poorly understood, even at a fundamental biophysical level. Recent advances in computational modeling offer promising opportunities to investigate gastric function. While experimental studies are indispensable, computational models offer a unique advantage: they enable the exploration of hypothetical physiological configurations and states that are difficult or impossible to replicate in the lab. This makes them powerful tools for efficiently testing and comparing competing biophysical hypotheses.

However, the evaluation of complex models across many different scenarios poses a significant challenge, as the computational cost can become prohibitively high. To address this, we introduce a framework for automated model evaluation in multi-query settings, encompassing parameter studies, sensitivity analysis, uncertainty quantification, and Bayesian parameter identification (i.e., model calibration). We demonstrate how even relatively simple methods—such as structured designs of experiments—can provide valuable insights and enable systematic comparisons of competing hypotheses. A key enabler of more advanced, large-scale studies is the use of surrogate modeling, which dramatically reduces computational costs and facilitates more sophisticated analyses such as global sensitivity analysis.

Using this framework, we for example show that replicating key features of gastric electromechanics—such as the extreme deformations observed experimentally in the fasted stomach—requires very specific conditions. In particular, our results highlight the need for a finely tuned coordination between contractions in the circumferential and longitudinal muscle layers. This emphasizes the critical coupling between electrical excitation and large-scale, nonlinear tissue deformation that underlies normal gastric function.

By integrating multi-query analysis techniques with advanced computational models—such as a patient-specific active-strain electromechanical model of the human stomach—we uncover novel insights into the biomechanics of gastric motility. This approach provides a powerful tool for deepening our understanding of both normal physiology and motility disorders. Ultimately, we envision this work as a bridge between computational modeling and clinical practice, paving the way for model-based diagnostics and treatment planning for disorders such as gastroparesis, GERD, and dyspepsia.

Introduction:
Biomechanical and electrochemical computational models of the stomach contraction have advanced significantly over the past two decades, progressing from single-cell simulations to whole-organ ones. More recently, chemo-electromechanical models have emerged, aiming to simulate the relationship between ion concentrations, electric potentials generated by the smooth muscles, and tissue contraction dynamics. The model presented here builds on these foundations by simulating slow wave generation and transmission across an anatomically inspired stomach scaffold, while independently visualizing smooth muscle contractions. A novel aspect of this work is the incorporation of a neural control mechanism that simulates the effect of neurotransmitters such as acetylcholine and nitric oxide on slow wave propagation and muscle contraction across the entire stomach.

Method:
The model couples a tridomain electrophysiological framework with a biomechanical spring-based model. The tridomain equations govern the propagation of slow waves through interstitial cells of Cajal, smooth muscle cells, and a fixed homogenous extracellular domain. Neural regulation pathways affecting specific ionic currents were incorporated based on recent physiological data, modulating excitatory and inhibitory effects through parameters such as Ano1 and NSCC channel conductances. The mechanical behavior was modeled by linking intracellular calcium concentration to tissue tension using Hill-type equations, and displacements were visualized through a damped harmonic oscillator analogy. The anatomical scaffold was derived from fitting CT images obtained from NIH’s SPARC database, and later processed into a finite element mesh in MATLAB, over which the model was solved using isotropic assumptions with Dirichlet and Neumann boundary conditions. Electrophysiological simulations were performed in MATLAB, while scaffold displacements were visualized in Python.

Results:
Simulation results demonstrated realistic propagation of slow waves originating at the gastric pacemaker region, traveling at physiological frequencies (~3 cycles per minute) towards the pylorus, with circumferential conduction velocities approximately double the longitudinal ones. The model also showed the expected frequency and amplitude modulation under neural stimulation: excitatory neural inputs increased wave amplitude by approximately 2.9 mV and raised contraction frequency to 4 cpm, while inhibitory inputs significantly reduced contractile tension by up to 29 kPa. Displacement simulations showed tissue deformations ranging from 0 mm at rest to approximately 2.9 mm during contraction, consistent with physiological observations. Validation was achieved through qualitative comparison with previous computational models of gastric motility, confirming correct slow wave directionality, frequency, and muscle deformation magnitude.

Conclusion and Future Work:

These results highlight the model's capability to simulate gastric motility with neural control on an anatomically inspired scaffold. However, in order for this model to serve as a platform for the design and in-silica testing of pharmaceutical drugs among other clinical applications, it requires further enhancements. Main aspects that will require future attention is the implementation of an anisotropic framework, a more accurate biomechanical model compared to the current Hill model and the incorporation of more neural control pathways.

The intestine plays a fundamental role in the digestive process, facilitating the transport and absorption of nutrients through complex motility patterns such as peristalsis and segmentation. Despite its importance, the precise mechanisms governing these movements remain only partially understood, making their mathematical modeling particularly challenging. Over the years, researchers have developed a variety of multiphysics and multiscale models to describe the intestine’s electrophysiology, mechanical behavior, and electromechanical interactions [1-2]. However, a crucial aspect that has been largely overlooked in existing models is the role of contact and self-contact during intestinal contractions.

In physiological conditions, different sections of the intestine may come into contact due to its natural motility. This phenomenon is further exacerbated in certain pathological conditions, where abnormal contractions or changes in tissue stiffness can lead to excessive compression or folding.This may lead to pathological conditions such as herniation or adherent syndromes causing obstructions. Moreover, during peristaltic contractions or segmentation, the inner surface of the intestine can collapse onto itself, leading to a form of internal self-contact that influences mechanical stress distribution and chyme transport. To the best of our knowledge, no existing model explicitly accounts for these contact phenomena in the simulation of intestinal motility.

To address this limitation, we propose a novel electromechanical model tailored for patient-specific simulations of intestinal motility, explicitly incorporating self-contact. Our approach builds upon the framework introduced in [1], utilizing the active strain approach to couple the electrophysiological and mechanical components of intestinal contractions. Additionally, we implement an unbiased contact formulation based on the Nitsche method [3], which eliminates the traditional master-slave hierarchy, ensuring a more robust and symmetric treatment of contact interactions. The proposed methodology is validated through benchmark tests before being applied to investigate self-contact phenomena in the duodenum and colon, providing new insights into the mechanical behavior of intestinal motility.

Keywords: gastrointestinal motility, unbiased contact method, electromechanics, patient-specific

References

[1] R. T. Djoumessi, P. Lenarda, A. Gizzi, S. Giusti, P. Alduini, and M. Paggi, In silico model of colon electromechanics for manometry prediction after laser tissue soldering. Comput. Methods Appl. Mech. Eng. (2024) 426: 116989.

[2] P. Du, S. Calder, T. R. Angeli, S. Sathar, N. Paskaranandavadivel, G. O'Grady, and L. K. Cheng, Progress in mathematical modeling of gastrointestinal slow wave abnormalities. Front. Physiol. (2018) 8:113

[3] R. Mlika, Y. Renard, and F. Chouly, An unbiased Nitsche’s formulation of large deformation frictional contact and self-contact. Comput. Methods Appl. Mech. Eng. (2017) 325: 265–288.

A properly functioning gastrointestinal (GI) tract plays a vital role in digestion, nutrient absorption, microbial homeostasis, and motility. GI pathologies can severely impair these functions, often necessitating surgical intervention involving resection, excision, or anastomosis, depending on the specific pathology. Post-surgical healing of GI tissues proceeds through the canonical phases of hemostasis, inflammation, proliferation, and remodeling. Even though, recent years have seen significant advances in electromechanical modeling [1] of GI tissues, and 3D bioprinting-based scaffolds have emerged to support both tissue repair and inflammation control, challenges persist in controlling and optimizing GI healing process.

Anastomotic leakage and dehiscence remain critical postoperative complications, associated with high morbidity and mortality. Although robust, clinically validated models exist for skin wound healing [2], analogous models for GI tissue are lacking due to the anatomical and physiological complexities of the GI environment. In recent developments, hydrogel-based biomaterials [3] are increasingly replacing traditional staples and sutures for GI tissue repair. However, the healing mechanisms associated with such materials also remain poorly understood. To the best of our knowledge, no existing model integrates self-healing and scaffold-based regeneration to elucidate the GI healing process.

We couple an electro-mechanical model of GI tissue with a scaffold-based healing framework to investigate both intrinsic and biomaterial-assisted regeneration. The model is used to optimize scaffold parameters for accelerated healing by modulating key biological markers such as cell density, cytokine concentration, collagen fibers alignment, and tissue contraction. The problem will be formulated as a dynamical system in which the control variables are the material and chemo-diffusive parameters of the 3D bioprinted construct evolving in time, and the target state is the restored GI physiological motility. Optimization algorithms will be used to minimize the loss function and find the best set of material parameters that the new biomaterial requires to restore the integrity of the host tissue. This study will enable a mechanistic understanding of scaffold-mediated healing, offering a computational tool for guiding biomaterial design in GI surgery.

References:

[1] Djoumessi RT, Lenarda P, Gizzi A, Giusti S, Alduini P, Paggi M. In silico model of colon electromechanics for manometry prediction after laser tissue soldering. Comput Methods Appl Mech Eng 2024;426.

[2] Sohutskay OP, Tepole AB,Voytik-Harbin SL. Mechanobiological wound model for improved design and evaluation of collagen dermal replacement scaffolds. Acta Biomater 2021;135.

[3] Su Y, Ju J, Shen C, Li Y, Yang W, Luo X, et al. In situ 3D bioprinted GDMA/Prussian blue nanozyme hydrogel with wet adhesion promotes macrophage phenotype modulation and intestinal defect repair. Mater Today Bio 2025;31:101636.

Idiopathic Pulmonary Fibrosis (IPF) is a disease characterized by the progressive formation of scar tissue in the lungs, leading to locally increased tissue stiffness and impaired respiratory function. To simulate lung behavior, micro-scale and personalizable organ-scale models have recently been developed [1,2,3]. However, integrating these models into a unified multi-scale model that can be personalized based on image data is computationally expensive. In our work, we propose a machine learning-based surrogate modeling framework that could be used to integrate the micro-scale model with the organ-scale model as well as to speed up the personalization procedure.

In this contribution, we investigate the properties of Hierarchical Deep-learning Neural Networks (HiDeNN) [4,5] and introduce several modifications to extend their capabilities [6]. Similar to classical Physics-Informed Neural Networks (PINNs), HiDeNN incorporates knowledge of the underlying physics. However, unlike standard PINNs, the solution is obtained as a linear combination of classical finite element shape functions. Due to their mesh-based structure, HiDeNN models require significantly fewer trainable parameters than fully connected neural networks. Additionally, individual weights and biases have a clear interpretation, which can be leveraged during training.

We compare two approaches to shape function construction: the interpolation-layer-based approach and the reference-element-based approach. In our analysis of model properties, we investigate the impact of different loss functions on training efficiency and solution accuracy, as well as the effect of various numerical integration methods for loss evaluation. Furthermore, we introduce two key extensions to the existing HiDeNN framework: rh-adaptivity, which employs a Jacobian-based criterion to guide adaptive mesh refinement, and a multi-level training strategy that leverages the model’s interpretability to enhance accuracy and computational efficiency. The framework's capabilities are demonstrated on 1D, 2D, and 3D test cases.

[1] Patte, C., Brillet, P.-Y., Fetita, C., Bernaudin, J.-F., Gille, T., Nunes, H., Chapelle, D., and Genet, M. (2022a). Estimation of regional pulmonary compliance in idiopathic pulmonary fibrosis based on personalized lung poromechanical modeling. Journal of Biomechanical Engineering, 144(9):091008.

[2] Patte, C., Genet, M., and Chapelle, D. (2022b). A quasi-static poromechanical model of the lungs. Biomechanics and Modeling in Mechanobiology, 21(2):527–551.

[3] Genet, M., Manoochehrtayebi, M., & Bel-Brunon, A. (2023, October). A micro-poro-mechanical model of the lung parenchyma. In SB 2023-48ème Congrès de la Société de Biomécanique.

[4] Zhang, L., Cheng, L., Li, H., Gao, J., Yu, C., Domel, R., Yang, Y., Tang, S., Liu, W.K., 2021. Hierarchical deep-learning neural networks: finite elements and beyond. Computational Mechanics 67, 207–230.

[5] Zhang, L., Lu, Y., Tang, S., Liu, W.K., 2022. HiDeNN-TD: reduced-order hierarchical deep learning neural networks. Computer Methods in Applied Mechanics and Engineering 389, 114414.

[6] Škardová, K., Daby-Seesaram, A., & Genet, M. (2024). Finite Element Neural Network Interpolation. Part I: Interpretable and Adaptive Discretization for Solving PDEs. arXiv preprint arXiv:2412.05719.

The remarkable properties of spider silk, including supercontraction - a significant shrinkage upon wetting - are deeply rooted in the molecular structure of its proteins. Understanding how the primary sequences of silk proteins influence macroscopic fiber behavior is essential for advancing biomimetic materials. In this work, we enhance a previously proposed microstructure-inspired multiscale model for predicting the spider silk supercontraction, by integrating new insights from data-driven modeling [1].

Specifically, by using Evolutionary Polynomial Regression (EPR), a machine learning technique combining genetic programming with symbolic regression, we investigate the relationships between the molecular composition of the silk main proteins and the fiber supercontraction response. Our analysis based on a recent multiscale experimental data across different silk types [2] identifies key sequence motifs: the repeat length of MaSp2 and the polyalanine regions of MaSp1 emerge as fundamental determinants of supercontraction dynamics. EPR generates interpretable mathematical relationships linking these sequence features to fiber contraction, which we interpret in mechanical terms. We propose that the polyalanine domains of MaSp1 regulate β-sheet misalignment, accommodating the shortening of softer regions during supercontraction. Meanwhile, the repeat length of MaSp2 governs the cross-linking interactions that stabilize amorphous chains in the dry state, while hydration disrupts them and triggers macroscopic fiber contraction.

Validation against experimental data from the Silkome database confirms the predictive capability of the proposed model, bridging molecular-scale protein structure with macroscopic fiber behavior. This work offers new insights into the molecular basis of supercontraction and provides a framework for designing biomimetic silk materials with tunable properties.

[1] Fazio, V., Malay, A. D., Numata, K., Pugno, N. M., & Puglisi, G. (2024). A Physically‐Based Machine Learning Approach Inspires an Analytical Model for Spider Silk Supercontraction. Advanced Functional Materials, 2420095.

[2] Arakawa, K., Kono, N., Malay, A. D., Tateishi, A., Ifuku, N., Masunaga, H., ... & Numata, K. (2022). 1000 spider silkomes: Linking sequences to silk physical properties. Science advances, 8(41), eabo6043.

There have been impressive advancements in the application of physics to the modelling of complex cardio-physiological systems, including the dynamics of the blood flow-pressure dependence in the vasculature connected to the heart. In principle this affords opportunities for deeper insight into the nature, cause and best treatment of cardio-physiological diseases. However, the corresponding mathematical models typically do not accommodate closed form solutions and entirely rely on numerical simulation procedures instead. This becomes problematic in clinical applications, where model calibration and patient specific parameter estimation are indispensable, calling for repeated forward simulations from the model as part of an iterative optimisation or sampling procedure at substantial computational costs. A potential workaround is to rely on surrogate models which approximate the simulations. The present work introduces an efficient method of predicting fluid dynamics in systemic arterial networks via physics-informed neural networks. In particular, we explore predictions associated with a 1D-fluid dynamics network model that accepts biophysical parameters as inputs, and as such functions as a surrogate model for numerical solvers of the fluid dynamics problem. The focus of our work lies on patient-specific modelling and requires a ventricular geometry profile as well as blood flow measurements in the ascending aorta which define the inlet boundary condition. Once fully trained, the machine learning model predicts blood flow and pressure waveforms in a fraction of the time taken by the numerical solver, allowing for fast uncertainty quantification of the parameters in the system given observational data using Markov chain Monte Carlo. Our inference framework is applied to magnetic resonance imaging (MRI) and magnetic resonance angiography (MRA) data from four patients diagnosed with double outlet right ventricle (DORV), a congenital heart defect where both the aorta and pulmonary artery originate from the right ventricle, instead of the left and right ventricle respectively. We show that our method provides an accurate non-invasive method of predicting blood pressure in the arteries surrounding the heart, accounting for uncertainty in inferred physiological parameters.

Introduction

Biomechanical stimuli are crucial in the progression of ascending thoracic aortic aneurysm (ataa) [1,2,3]. Currently, biomechanical stimuli are typically obtained in silico through computational models. However, debate remains regarding computational fluid dynamics (cfd) and computational solid mechanics (csm) simulations due to (i) modeling assumptions and incomplete/uncertain input data derived from clinical sources and (ii) the complex interplay between hemodynamic and structural stresses, which is often overlooked. This study introduces a semi-automatic fluid-structure interaction (fsi) framework that integrates the identification of patient-specific aortic wall material properties, aiming at minimizing a primary source of uncertainty while enabling a comprehensive assessment of aortic biomechanics.

Methods

Aortic geometries, including the valve orifice plane, aortic root, and supra-aortic vessels, are reconstructed at peak systole and at end-diastole from retrospectively-gated computed tomography angiography (cta) [2]. Patient-specific transaortic jet velocity measured by Doppler echocardiography and pressure data obtained through brachial cuff measurements are used as input data. The vessel wall is modeled as a nonlinear isotropic hyperelastic Neo-Hookean solid with uniform thickness. To find patient-specific material properties, a simple iterative approach is adopted, using the ascending aorta (aao) measured systolic volume as target value for the calibration. After obtaining the diastolic tensional state through prestress simulations [4], the patient-specific systolic blood pressure is applied to the aortic wall in csm simulations. Young’s modulus is varied iteratively bisecting the initial interval 0.8-8.0 MPa [3] until the simulated aao systolic volume matches the measured target value within a 1% error. Following a prestress simulation using the previously estimated Young’s modulus, fsi simulations are conducted with a two-way arbitrary Lagrangian-Eulerian (ale) formulation. Homogeneous Dirichlet conditions are applied at the ring-shaped solid domain boundary sections in terms of null displacement. Doppler-derived flow rates are imposed in terms of flat velocity profile at the fluid inflow section. Zero-dimensional (0d) models are coupled to the 3d fluid domain outflow sections. The 0d parameters are tuned to match patient-specific cuff pressures in coupled 0d and one-dimensional (1d) simulations. All simulations are performed using the finite element-based solver SimVascular.

Results and discussion

For a representative patient, the aao peak systolic volume measured from cta was 107.60 cm³, leading to an estimated Young’s modulus of 2.15 MPa. The fsi simulation was able to capture the flow impingement region located in the anteromedial area of the bulge, where the highest values of time-averaged wall shear stress (tawss) were recorded, along with the greatest variation in wss contraction and expansion action on the luminal surface along the cardiac cycle.

By reconstructing the valve orifice, the fsi framework allows to simulate both tricuspid and bicuspid valve morphologies. This framework enables the exploration of the synergistic action of hemodynamic and structural stress, potentially revealing distinct or combined effects of these biomechanical stimuli on ataa progression.

References

1. De Nisco et al., Med Eng Phys, 2020.

2. Pasta et al., Ann Thorac Surg, 2020.

3. Trabelsi et al., Ann Biomed Eng, 2016.

4. Bäumler et al., Biomech Model Mechanobiol, 2020.

Accurately capturing the dynamics of blood flow, pressure distribution, and structural deformation within the human heart is essential for advancing our understanding of cardiovascular physiology and for improving the diagnosis and monitoring of heart disease. However, direct in-vivo measurements remain challenging: they are often invasive, limited in spatial and temporal resolution, and constrained by technical and ethical considerations. In this context, patient-specific digital twins (i.e., computational models that can reproduce individual cardiovascular anatomy and function) have emerged as a powerful tool for simulating the cardiac cycle with high fidelity.

In this work, we employ our in-house, multi-GPU simulation framework [1], designed to capture the full complexity of cardiac mechanics through a tightly integrated Fluid–Structure–Electrophysiology Interaction (FSEI) formulation. This framework incorporates a biophysically detailed electrophysiology model to simulate myocardial activation via a physiologically-informed conduction system. It also features anatomically realistic mitral and aortic valves, enabling accurate simulation of valve dynamics throughout the cardiac cycle.

A key component of our approach is the integration of patient-specific anatomical data, typically acquired at discrete time points during the cardiac cycle. These data are assimilated into the simulation to personalize the digital twin, ensuring that it reflects the subject’s unique geometry and boundary conditions. To validate this approach and evaluate its sensitivity to physiological variability, we use the full FSEI model as a reference, allowing us to assess both the realism and robustness of the personalized simulations.

To probe the complexity of intraventricular flow, we carry out a detailed Lagrangian analysis based on passive tracer dynamics. By computing statistical quantities such as structure functions, flatness, and temporal correlation functions, we characterize the multiscale features of turbulence within the left heart.
Our results show that both ventricle and aorta are marked by strong intermittency and long-lasting flow structures. We further explore how these flow characteristics are influenced by physiological parameters, including heart rate and aortic valve stiffness. Specifically, we find that a lower heart rate enhances flow intermittency by allowing coherent structures to develop and persist, while increased valve stiffness amplifies extreme fluctuations.

Overall, this study demonstrates the effectiveness of Lagrangian diagnostics in capturing the spatial and temporal complexity of cardiac flows. Our findings contribute to a deeper understanding of cardiovascular hemodynamics and highlight the potential of combining high-performance computing with clinical data to develop predictive, patient-specific digital twins of the heart.

This research was supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme Smart-TURB (Grant Agreement No. 882340), and by the project MUR-FARE R2045J8XAW.

[1] Viola et al., Scientific Reports, 13(1) (2023).

We examine how gravity affects bicuspid valves designed to prevent backflow in lymphatic vessels—an area with limited research. The study focuses on environments with varying gravitational conditions, such as microgravity during space missions and parabolic flights, or hypergravity during fighter flights. Computer simulations were conducted while varying both gravity intensity and valve length to verify valve performance. Results show that valve effectiveness is influenced by both gravity and valve length, with longer valves performing better—though this is moderated by gravitational force. These findings offer valuable insights into valve efficiency in biofluid pumping under non-standard Earth gravity conditions.

Aortic valve replacement is a surgical procedure to treat aortic valve diseases, such as stenosis and regurgitation. Bio-prosthetic valves, usually made from bovine or porcine pericardial tissue, mimic the natural flow dynamics of the native aortic valve and reduce the need for long-term anticoagulant therapy. However, the positioning of the stented frame relative to the native aortic annulus plays a critical role in determining the resulting blood flow patterns [1]. Specifically, the supra-annular configuration, characterized by positioning the valve ring superior to the native annulus while extending leaflets into the Valsalva sinuses, seeks to attenuate flow obstruction by circumventing interference with the native annulus. Conversely, the intra-annular approach, which closely mimics the physiological arrangement of the aortic valve, necessitates implantation of a smaller prosthesis with a corresponding reduction in valve effective orifice area, thereby potentially reducing hemodynamic performance [2]. This work focuses on the influence of deployment locations of bio-prosthetic aortic valves by examining blood flow characteristics in both intra-annular and supra-annular configurations. The problem is investigated numerically by solving fluid-structure interaction, incorporating the interplay between blood flow and the flexible valve leaflets within a realistic anatomy of the left heart. We select a small-sized heart, reproducing the scenario where patients with small aortic diameters suggest the surgeon to evaluate alternative surgical strategies. In addition, we consider a bioprosthetic valve model that resembles the LivaNova Crown PRT, specifically chosen because, unlike most prosthetic valves, it can be implanted in both configurations. This unique feature allows us to perform a controlled, high-fidelity numerical comparison within the same virtual patient, ensuring identical anatomical and functional conditions and isolating the effects of deployment strategy. The dynamic and transitional nature of hemodynamics is captured through direct solution of the incompressible Navier-Stokes equations using a staggered finite differences approach. Immersed boundary techniques are employed to address the large valve deformations. Structural mechanics is based on the Fedosov interaction potential method to model the behaviors of biological tissues [3-4]. The numerical model is used to compare hemolysis, wall shear stress, transvalvular pressure drop, and valve orifice area across the two prothesic deployment configurations.
This project has received funding from the European Research Council (ERC) under the European Union’s Horizon Europe research and innovation program (Grant No. 101039657, CARDIOTRIALS to F.V.). This work has received partial funding from the project MUR-FARE R2045J8XAW CUPE83C22005500001 (P.I. Luca Biferale).
[1] Vahidkhah & Azadani, Journal of biomechanics 58, 114-122 (2017).
[2] Kim et al., Interactive cardiovascular and thoracic surgery 28(1), 58-64 (2019).
[3] Viola et al., Scientific reports 13(1), 8230 (2023).
[4] Viola et al., Physical Review Fluids 8(10), 100502 (2023).

INTRODUCTION

Shoulder injuries, particularly those affecting the rotator cuff (RC), are common and result from repetitive movements, excessive strain, or trauma, leading to pain, muscle weakness, and reduced joint mobility. Rehabilitation, especially post-surgical, is essential to restore the range of motion (ROM) and prevent complications. However, recovery assessment is often based on subjective medical evaluations, limiting accuracy. While motion capture (MOCAP) systems offer more objective assessments, they are expensive and not easily accessible for routine clinical monitoring. An accessible system is needed to track shoulder movements during rehabilitation, enhancing therapy programs with real-time feedback.

METHODS

This study was funded by the European Union - Next Generation EU - PNRR M6C2 - Investment 2.1 Enhancement and strengthening of biomedical research in the National Health Service (Project No. PNRR-MAD-2022-12376080 - CUP: F83C22002450001). The study developed a wearable sensor based on fiber Bragg gratings (FBG) and 3D printing to monitor shoulder movements objectively and continuously. Made of thermoplastic polyurethane (TPU), known for its flexibility and durability, the sensor incorporates an embedded FBG sensitive to strain and temperature variations. The TPU substrate design, shaped like a "dog bone," optimizes strain transmission in the sensor's central section.

Two stretchable anchoring bands, integrated during printing, ensure stability and ease of use.

Tests evaluated strain sensitivity and hysteresis error using a tensile testing machine and temperature sensitivity through a laboratory oven. A preliminary test on a healthy volunteer assessed sensor performance in monitoring shoulder movements at different angles (0-30°; 0°-60°; 0°-90°) and speeds (0°-90° in 3s and 6s).

RESULTS

The sensor exhibited an average strain sensitivity (S_ε) of 1.45 nm/mε and a temperature sensitivity (S_T) of 0.02 nm/°C, slightly higher than a bare FBG due to TPU's thermal expansion. However, the hysteresis error was high (51%), indicating a nonlinear response under dynamic conditions.

During preliminary tests on a healthy volunteer, the sensor detected shoulder movements in the sagittal plane, with output variations (Δλ_B) proportional to ROM (0.22 nm at 30°, 0.46 nm at 60°, 0.77 nm at 90°) and speed (6-second cycles: mean Δλ_B of 0.784 nm ± 0.028 nm; 3-second cycles: mean Δλ_B of 0.762 nm ± 0.029 nm). These findings demonstrate the sensor's potential for real-time monitoring and clinical applications.

CONCLUSIONS

The developed sensor combines FBG and 3D printing to monitor shoulder movements during rehabilitation, offering high sensitivity and an ergonomic design. Preliminary results are promising but highlight limitations, such as high hysteresis error and TPU's thermal influence.

Future developments aim to improve linearity, reduce thermal effects, and lower hysteresis, as well as test the device on a larger sample. Additionally, sensor evaluations will be extended to shoulder movements in the frontal and scapular planes to provide a more comprehensive analysis in real-world conditions.

This sensor represents an affordable, portable, and non-invasive solution with the potential to revolutionize rotator cuff rehabilitation and other medical and sports applications, providing objective data to optimize therapy pathways and improve functional outcomes.

INTRODUCTION: Telerehabilitation represents an innovative solution for remotely monitoring and managing physical rehabilitation, enabling the execution and supervision of exercises in home or non-clinical settings. This approach has the potential to enhance the overall quality of rehabilitation care, promoting an inclusive strategy that addresses challenges related to accessibility and the sustainability of healthcare services. Among the various technologies explored, thermal cameras, when integrated with computer vision and artificial intelligence (AI) algorithms, offer significant advantages. They enable non-invasive monitoring and the detection of thermal variations associated with muscle movement and physical activity while also addressing privacy concerns, an essential aspect for home-based applications.
This study aims to evaluate the performance of an integrated system based on AI algorithms and thermal cameras for the automatic recognition of upper limb rehabilitation movements.

METHODS: This work was funded by the European Union - Next Generation EU - PNRR M6C2 - Investment 2.1 Enhancement and Strengthening of Biomedical Research in the National Health Service (Project No. PNRR-MAD-2022-12376080 - CUP: F83C22002450001). A total of 20 healthy volunteers were recruited. Data were acquired using a thermal camera (SEEK THERMAL Compact Pro) connected to an Android smartphone, positioned at a fixed distance from the subjects. The experimental protocol included ten shoulder rehabilitation exercises, each repeated an average of six times in three different scenarios: (1) Flexion/Extension, (2) Abduction/Adduction, (3) Scapular plane elevation, (4) External rotation, (5) External rotation with shoulder at 90° abduction, (6) Military press in standing position, (7) Wall slide in standing position, (8) Towel slide, (9) Forward bow, and (10) Pendulum. After a preprocessing phase, the recorded videos were processed through a pipeline that included frame extraction and the identification of 17 anatomical key points. These data were subsequently used to train (70% of the dataset) and validate (30% of the dataset) different AI models, including both machine learning (ML) and deep learning (DL) ones, for the automatic classification of rehabilitation exercise types.

RESULTS: The proposed computer vision approach demonstrated excellent performance on the test data. The findings show the superior performance of the DL model, achieving 95% accuracy in classifying the different exercise classes and an average Area Under the ROC curve of 0.97. The model was able to autonomously extract both spatial and temporal features from the input data, enabling it to capture more complex patterns and relationships within the data. The classification accuracies for all exercises ranged from 74% to 100%, underscoring the robustness of the model even when subjects performed the exercises in various scenarios.

CONCLUSIONS: The results demonstrate the effectiveness of the model in recognizing shoulder rehabilitation exercises. This study highlights the potential of thermography-based telerehabilitation as an innovative tool for remote monitoring. Such an approach can support both physiotherapists and patients in maintaining rehabilitation programs at home, reducing the need for in-person visits and enabling continuous supervision. Further studies should focus on evaluating the performance of such approach on data collected from patients with musculoskeletal disorders, and incorporating objective assessment of the executed exercises.

Purpose
A recent pilot study (Raso et al., J Exp Orthop, 2024) highlighted the potential of the NAO humanoid robot in guiding shoulder rehabilitation exercises, demonstrating promising results in improving exercise consistency and patient engagement. However, that work relied heavily on an external optical motion‐capture laboratory to track shoulder kinematics, creating barriers to widespread clinical and home‐based adoption. Building on those findings, this new study explores an AI‐driven approach to embed motion‐tracking capability directly within the NAO robot using its integrated camera. Our overarching goal is to remove the reliance on specialized motion‐capture labs and thereby extend the possible reach of robotic rehabilitation—whether in smaller outpatient clinics or in patients’ homes.

Methods
We modified our existing NAO‐based rehabilitation protocol, previously validated with an external motion‐capture system, to incorporate a custom AI module for onboard motion capture. In this updated configuration, NAO uses its camera feed to perform live pose estimation, focusing on key landmarks around the shoulder complex. A convolutional neural network (CNN) processes these video frames in real time, estimating the user’s joint angles and range of motion (ROM) without external sensors. The robot then provides adaptive, step‐by‐step verbal and gestural cues to guide flexion–extension, external rotation, and internal rotation exercises at varying speeds. Preliminary data collection involved a small sample of healthy individuals replicating the earlier lab‐based exercises. By comparing NAO’s internal AI‐generated kinematics to “gold standard” motion‐capture metrics, we qualitatively assessed accuracy, ease of use, and potential for clinical integration.

Results
Preliminary analyses suggest that NAO’s onboard AI system can capture and estimate shoulder ROM with moderate fidelity relative to the optical laboratory standard—particularly within mid‐range movement arcs crucial for early to mid‐stage rehabilitation. While greater deviations arose at the extremes of the ROM, the absolute mean errors observed in pilot sessions remain within clinically acceptable thresholds for guiding exercise performance. Participants reported a high level of confidence in NAO’s real‐time feedback and found the single‐unit robotic system easier to set up than traditional motion‐capture equipment. These early results mirror the positive engagement documented in our earlier study, indicating that the robot‐patient interaction is maintained without needing an external camera system.

Conclusions
By integrating an AI‐powered motion‐tracking module directly into NAO, we build on the success of our prior study and advance toward a more accessible form of robot‐assisted therapy for shoulder rehabilitation. Although further refinements and larger clinical trials are needed—especially to verify accuracy in individuals with shoulder pathologies—our findings underscore the potential of a “one‐system” solution to deliver robust rehabilitation guidance in non‐specialized settings. If ongoing development continues to validate these early results, clinicians could deploy NAO in small outpatient clinics or patients’ homes, reducing the need for fully equipped motion‐analysis laboratories. This shift toward accessible AI‐driven robotics could significantly expand the scope of shoulder rehabilitation, improving patient adherence and long‐term outcomes while decreasing the burden on clinical facilities.

Motile cilia are slender, flexible hair-like organelles that are used by many cells to move and manipulate the fluids that surround them. Their effectiveness for this task is evidenced by their usage across eukaryotic life, from unicellular organisms that use cilia for propulsion, up to complex, multicellular organisms including humans, where cilia play a fundamental transport role in vital organs. Cilia are essential to the protection of our lungs, allowing for the effective clearance of the mucus that provides a mechanical barrier to pathogens and pollutants.

Modelling cilia motion and the flow fields that they generate has been an active topic of research in biofluiddynamics since the inception of the field. Early work focused on understanding individual cilium motion using drag-based models, such as resistive force theory, or describing the flows generated by cilia arrays by imposing velocities on ciliated boundaries.

In this talk, I will describe our recent work that takes advantage of the growth of computing power to bring together these two modelling paradigms by considering large-scale arrays of coupled cilia, connecting resulting flow fields with collective motion. Our approach, stemming from the Lagrangian Mechanics of Active Systems (Solovev, A. and Friedrich, B.M., 2021. Lagrangian mechanics of active systems. The European Physical Journal E, 44, pp.1-15), allows for our computations to link directly with experimentally measured cilia beats, while also reducing the number of dynamic degrees of freedom and providing the model flexibility needed to probe additional physical characteristics, such as cilium elasticity. In addition to describing this framework, I will present recent results illustrating the emergent metachronal waves given by the model and how they provide new insights into cilia driven flows. While these results are linked primarily to studies involving model organisms, I will describe some perspectives and extensions of the model to study mucociliary clearance in the lung.

Mathematical modeling of the respiratory system provides a structured framework for analyzing various processes involved in breathing. This approach may be particularly relevant to improve diagnosis in the context of lungs pathologies such as asthma, COPD, fibrosis or emphysema where spontaneous ventilation is impaired. In these situations, patients are admitted to intensive care units where artificial ventilation assists them with their breathing effort by injecting air through the trachea, leading to possible lung injuries. To describe the ventilation process and understand how the lungs react to these large pressure variations, we propose a dynamic continuum mechanics model describing the lung deformations as those of a hyperelastic material. We consider the non-linear constitutive law developped in [1] in the steady state setting and generalize it to an unsteady setting. Furthermore, we couple it to a 0D representation of the bronchial tree, viewed as a resistive diadic tree as in [2]. We choose to preserve the equivalent resistance of the tree while reducing the total number of generations, in an effort to lower computational cost. Several terminal subdomains are then defined, representing the region of influence of a path through the bronchial tree. This coupling between the air flow through the bronchial generations and the lung parenchyma deformations adds a pressure gradient in the elastic equations representing the action of the air flow on the elastic media. This additional term can be written as a non-local non-linear dissipative term which depends only on the parenchyma deformations. The resulting equations can be viewed as a poromechanical model describing the air-parenchyma interaction and satisfying energy estimates. To discretize in time the resulting system, we consider a Newmark scheme with a finite element approximation for the space discretization. Due to the non-local term, the associated finite element matrix is a dense matrix obtained by the product of two time-dependent matrices, greatly increasing computation and memory costs, especially as the system is solved by a Newton algorithm. To overcome this limitation, we propose different numerical strategies as well as a local approximation of the resistive term. Both non-local and local models are first compared on a simplified 2D case. In particular, we focus on their ability to dissipate energy and reproduce the variations of the lung volumes. Next, normal breathing scenarios are simulated for a 3D realistic lung geometry considering some boundary conditions modeling the impact of surrounding elements such as the rib cage or the diaphragm. The proposed coupled models shall be further used in pathological situations in particular to explore possible lung injuries for patients under artificial ventilation.

[1] Patte C, Genet M, Chapelle D. A quasi-static poromechanical model of the lungs. Biomech Model Mechanobiol. 2022

[2] Pozin N, Montesantos S, Katz I, Pichelin M, Vignon-Clementel I, Grandmont C. A tree-parenchyma coupled model for lung ventilation simulation. Int J Numer Meth Biomed Engng. 2017

Mechanics and, more specifically, stress fields possibly play a crucial role in the development of pulmonary fibrosis. This work aims to provide clinicians with diagnostic and prognostic tools based on mechanical simulation. Personalisation of these tools is critical for clinical pertinence, thus requiring numerical techniques for real-time estimation of patient-specific mechanical parameters. This study presents a hybrid approach [1] that integrates the Proper Generalised Decomposition (PGD) [2] with deep learning techniques to develop real-time diagnostic and prognostic tools for clinicians, particularly in assessing compliance fields in fibrotic lungs.
The proposed method mitigates the curse of dimensionality inherent in parametric systems by employing a tensor decomposition. The building brick [3] of the proposed framework is based on the concept of HiDeNN [4]. Indeed, each mode of the tensor decomposition is represented by a sparse neural network with constrained weights and biases to replicate standard Finite Element Method (FEM) shape functions. This constraint enhances model interpretability and facilitates transfer learning, significantly accelerating the training process. Moreover, the model's architecture is directly determined by the mesh and the interpolation order, eliminating arbitrary choices and allowing mesh adaptation during the training stage.
Similarly to PINNs, the physics of the problem is incorporated into the loss function during unsupervised training. The training process involves solving a minimisation problem, similar to classical model reduction. However, automatic differentiation within the neural network framework allows for greater flexibility in addressing non-linearities, particularly when linearisation is difficult. The framework, therefore, offers a flexible framework for surrogate modelling in non-linear mechanics, as demonstrated through 1D, 2D and 3D benchmark problems that validate its robustness against analytical and numerical reference solutions.
As mentioned, the surrogate model is meant to be a digital twin of the lungs. Its parametrisation must, therefore, account for the shape variability of lungs from one patient to another. To that aim, a shape registration pipeline from CT images has been developed, from which a library of shapes can be obtained and parametrised. Each patient shape parameters can then be fed into the surrogate model to get a patient-specific solution.
References
[1] Daby-Seesaram, A., Skardova, K., & Genet, M. Finite Element Neural Network Interpolation: Hybridisation with the Proper Generalised Decomposition for surrogate modelling. Submitted,
[2] Chinesta, F., Ladeveze, P., & Cueto, E. (2011). A Short Review on Model Order Reduction Based on Proper Generalized Decomposition. Archives of Computational Methods in Engineering, 18(4), 395–404. https://doi.org/10.1007/s11831-011-9064-7
[3] Skardova, K., Daby-Seesaram, A., & Genet, M. Finite Element Neural Network Interpolation: Interpretable Neural Networks for Solving PDEs. Submitted,
[4] Zhang, L., Cheng, L., Li, H., Gao, J., Yu, C., Domel, R., Yang, Y., Tang, S., & Liu, W. K. (2021). Hierarchical deep-learning neural networks: Finite elements and beyond. Computational Mechanics, 67(1), 207–230. https://doi.org/10.1007/s00466-020-01928-9

Mechanical ventilation is a life-saving therapy for conditions such as acute respiratory distress syndrome (ARDS), requiring careful balancing of breathing effort reduction, gas exchange improvement, and prevention of alveolar overdistension. Achieving optimal settings is challenging due to inter-subject variability, frequent comorbidities, and competing therapy objectives. Computational lung models integrating patient-specific anatomy and respiratory mechanics offer promising personalized strategies. However, despite the emergence of novel models, validation against clinically relevant experimental data remains scarce, hindering their path toward clinical application.

This work presents a digital-twin framework that couples a three-dimensional nonlinear finite-element model of lung parenchyma with a zero-dimensional airway-tree flow network. We validate the framework using porcine data from five ARDS-model subjects. The lung geometries were reconstructed from end-expiratory CT scans. Airway geometries were segmented and extended through a generative branching algorithm to capture sub-voxel airway generations. The lung parenchyma was segmented into high-fidelity meshes, with regional porosity fields derived from CT Hounsfield units to define spatially variable permeability distributions.

Two constitutive models, a Yeoh-style hyperelastic model and an exponential model, were implemented and calibrated under volume-controlled ventilation to ensure physiologically plausible tissue strains and pressures. The incorporation of pre-strain into our simulations allows for the introduction of a baseline pressure, akin to PEEP in volume-controlled ventilation. These models were then applied to simulate two clinically relevant strategies: the low-tidal-volume ARDSnet protocol and airway-pressure-release ventilation (APRV). Simulated global signals of airway pressure, flow, and volume were generated, closely resembling typical clinical ventilator outputs. Simultaneously, the digital-twin framework resolved the regional mechanical response of lung parenchyma, generating detailed distributions of deformation, strain, stress, and other continuum-mechanics-derived indicators.

Model evaluation was performed at both global and regional scales. On a global scale, simulated airway pressure, volume, and flow were compared against experimental recordings for both volume- and pressure-controlled ventilation. Regionally, end-inspiratory aeration distributions were compared to those derived from CT images. Deformation fields were compared to image-registration-derived deformation maps, enabling assessment of differences between the image-registration reference standard and predictions from the models under varying conditions, including ventilation modes and constitutive models. Tidal ventilation distribution metrics were further compared to regional ventilation measurements from imaging sequences.

This integrated evaluation highlights the strengths and limitations of the current framework. Globally, the coupled poromechanical-airway model successfully reproduced key features of respiratory mechanics across ventilation modes. Regionally, discrepancies in predicted aeration and deformation underscore areas for model improvement, particularly in regions with pronounced heterogeneity and boundary interfaces, where airway branching variability influences local ventilation.

These insights will guide future refinements, including incorporating viscoelastic and surface-tension effects, enhanced heterogeneity mapping, and improved chest-wall and diaphragm interactions.

The Growth and Remodeling (G&R) mechanisms in the arterial wall have been extensively investigated over the past two decades, with particular focus on their involvement in disease progression such as aneurysms. The arterial wall's microstructure is composed of various cellular and extracellular matrix elements. From a mechanical standpoint, collagen and elastin represent the most essential extracellular constituents. Collagen fibers provide structural reinforcement to the tissue. Additionally, as with most soft biological tissues, the arterial wall preserves a specific preferred mechanical state (mechanical homeostasis) [1,2]. Although computational constrained mixture models of arterial G&R have significantly enhanced our comprehension of this complex phenomenon, their clinical use—such as in computer-assisted treatment or surgical planning—remains unfulfilled. When the tissue deviates from this preferred mechanical state, G&R processes, involving modifications in mass and internal organization, work to reestablish homeostasis. The G&R mechanism alters the collagen fiber orientation distribution in the tissue, leading to time-dependent degradation and deposition of fibers. Consequently, incorporating G&R behavior into models is fundamental for understanding the arterial wall’s mechanical response and aneurysm formation. Building on previous studies where the authors introduced a statistical framework offering analytical insights into the evolution of collagen fiber alignment in soft tissues during G&R under uniaxial loads [3], the present work aims to apply that theoretical framework to explore G&R in both healthy and aneurysmatic arterial walls. The arterial wall is modeled as a mixture of multiple constituents deforming collectively. By employing the homogenized constrained mixture theory [3] along with the principle of mechanical homeostasis [1], a probabilistic law for fiber-mass evolution is formulated and solved numerically. As collagen fibers are not oriented uniformly but follow a statistical distribution, this formulation includes a probability distribution of fiber orientations [4], assumed to follow a Von Mises-type distribution. To facilitate analysis, the Von Mises distribution is taken to have the same mean and variance as the fiber mass density distribution. The homogenized constrained mixture model from [3] is then used to analyze G&R in both healthy and aneurysmal arterial walls. A novel set of differential equations describing elastic and inelastic deformations in the circumferential, radial, and longitudinal directions is derived. Once the evolution of the fiber probability density is known, the remodeling behavior of the tissue in both healthy and pathological states can be evaluated. Numerical simulations and comparison with experimental data highlight the method’s effectiveness. The proposed method offers a promising strategy for modeling G&R processes in arterial tissue under various conditions. Preliminary simulations show strong correlation with experimental data, and a qualitative comparison with existing G&R models is also presented.

Introduction

The treatment of cardiovascular disease places a substantial burden on global healthcare systems [1]. Vascular pathologies such as aortic dissection, stroke, and aneurysm can lead to the rupture of the vascular wall, often with fatal consequences. Predicting and preventing such ruptures requires a deep understanding of the mechanics governing fracture behavior in vascular tissue [2]. Given its hierarchical structure, vascular tissue exhibits complex mechanical properties, necessitating robust and efficient numerical tools for identifying its fracture mechanics. While the elastin network facilitates vessel wall recoil, collagen fibers play a crucial role in non-linear viscoelasticity, supporting large deformations and providing both stiffness and toughness. The orientation of these fibers varies within the vascular wall and across different aortic locations, influencing its macroscopic mechanical properties.

Material and Methods

This study focuses on characterizing the fracture properties of the aortic media, a vessel wall layer where collagen fibers are primarily aligned in the circumferential direction. Our previously established in-vitro symmetry-constrained Compact Tension (symconCT) test [3] serves as ground truth data for Finite Element Modeling (FEM). In an attempt to overcome drawbacks of our more classical fracture modeling approaches (cohesive zones [4], extended FEM [5]), we now explored phase-field approaches [6,7].

The vessel wall shows anisotropic mechanical behaviour [8], a property which relation to fracture propagation is not well understood. We therefore described vessel wall bulk properties with both, the isotropic Yeoh constitutive model and the anisotropic Gasser-Ogden-Holzapfel (GOH) model and explored its implications. As viscoelasticity has a paramount implication in tissue fracture [9], a five-element Maxwell description captured the time-dependent tissue properties. Similarly to the description of the bulk material properties, isotropic and anisotropic phase field description have been tested. All our models have been realized within FEAP (Univ. of California at Berkeley, US).

Results and conclusions

Phase field modelling allows to deal with most complexity of vascular tissue fracture, requires however significant numerical resources. An anisotropic description of the bulk properties of the normal vessel had a minor implication on the fracture propagation direction. In highly diseased specimens from the aneurysmatic aorta, fracture is often diverted along the circumferential direction [10], an experimental observation that could only be captured by an anisotropic phase field description, and which was challenged by sever FEM mesh distortion.

References

1. (WHO), "W.H.O.: Cardiovascular Diseases (CVDs)"

2. Timmis, A et al., Eur. Heart J. 41, 12–85 (2020)

3. Alloisio, M et al., Acta Biomaterialia 167, 147-157 (2023)

4. Alloisio, M et al., Acta Biomaterialia 167, 158-170 (2023)

5. Miller, et al, Computational Mechanics 73 (6), 1421-1438 (2024)

6. Teichtmeister S et al., Int. J. of Non-Lin. Mech. 97, 1-21 (2017)

7. Aldakheel F., Leibniz Universität Hannover – Habilitation: https://doi.org/10.15488/11367 (2021)

8. Gasser TC, 1st edn. Springer https://doi.org/10.1007/978-3-030 (2021)

9. Forsell C, et al., Journal of biomechanics 44 (1), 45-51(2011)

10. Alloisio M, et al., Scientific Reports 15 (1), 667 (2025)

Introduction

The human aorta undergoes lifelong morphological changes driven by mechanical and biological factors. Aging and hypertension, in particular, increase aortic tortuosity, disrupting flow and wall stresses and raising the risk of cardiovascular complications. A key driver of these changes is the progressive loss of axial pre-stretch, a crucial factor for maintaining biomechanical stability. While previous studies have linked reduced axial pre-stretch to aortic elongation and altered mechanical behavior, its precise role in growth and remodeling (G&R) and its potential to induce pathological deformations, such as aortic buckling, remain poorly understood. This study aims to elucidate the influence of axial pre-stretch on aortic morphology evolution using a patient-specific finite element simulation.

Method

A computational model of the human ascending aorta was developed based on the constrained mixture theory, accounting for the distinct mechanical contributions of elastin, collagen, and smooth muscle cells. The model captures the tissue’s capacity to restore homeostatic stress via biological adaptation mechanisms such as mass growth and structural remodeling. It was implemented in Abaqus through a UMAT subroutine to simulate the G&R of the ascending aorta.

Using age-related physiological axial loading conditions, we considered two cases to evaluate effects of axial pre-stretch on the G&R of the ascending aorta in hypertensive patients. In the first case, representing a 70-year-old patient, the model was subjected only to 80 mmHg without axial pre-stretch. In the second case, corresponding to a 40-year-old patient, the same pressure was applied with 20% axial pre-stretch. After computing the pre-stress fields, both models were subjected to a 40% increase in pressure to simulate hypertensive conditions.

Results

The simulations revealed distinct remodeling patterns between the two scenarios. In both cases, the model predicted increases in aortic diameter and wall thickness, consistent with clinical observations of hypertension-related remodeling. However, the presence or absence of axial pre-stretch had a significant impact on aortic morphology and mechanical stability.

In the model without axial pre-stretch, the ascending aorta exhibited localized kinking along the inner curvature under elevated pressure, indicative of pathological deformation and reduced structural integrity. In contrast, the model with 20% axial pre-stretch preserved a smoother geometry, even following a 40% increase in blood pressure.

An analysis of stress evolution during G&R revealed a gradual reduction in axial stress over time in both models. This effect was more pronounced in the non-pre-stretched aorta, where the loss of axial tension led to localized pathological deformations. These findings are consistent with prior studies suggesting that buckling pressure decreases as axial pre-stretch diminishes. Our results confirm that insufficient axial tension compromises the mechanical stability of the aortic wall, increasing the risk of tortuous geometries.

Overall, the results underscore the critical biomechanical role of axial pre-stretch in maintaining aortic structure under hypertensive loading. The loss of axial tension appears to drive maladaptive remodeling and may serve as a key factor in the onset of tortuosity and buckling in hypertensive patients.

Pelvic Organ Prolapse (POP) is a prevalent condition among women globally, arising from weakened pelvic floor structures. Traditional surgical treatments frequently utilize synthetic meshes for reinforcement; however, these materials can cause complications such as erosion and chronic pain. Consequently, biodegradable mesh implants have emerged as promising alternatives due to their improved mechanical biocompatibility. This study investigates the mechanical properties of biodegradable polycaprolactone (PCL) meshes through experimental analyses and computational modeling, aiming to enhance their efficacy for POP repair.
The primary objective of the study is to evaluate the mechanical behavior of biodegradable PCL meshes compared to native vaginal tissue. This involves both experimental testing and numerical simulations. A key focus is assessing how different mesh geometries influence mechanical performance and validating computational models to enhance predictive accuracy for future POP repair applications.
In the methodology, biodegradable PCL meshes with quadratic and cross-shaped geometries were fabricated via melt electrowriting (MEW), with a filament thickness of 240 µm. Uniaxial tensile tests and ball burst tests were conducted on these meshes, alongside comparative tests on sow vaginal tissue. Numerical simulations using Abaqus/Explicit software were developed to replicate ball burst tests, allowing an evaluation of tissue-mesh interaction. Validation of these computational models was performed by comparing them with experimental data.
Results from ball burst testing demonstrated that biodegradable PCL meshes increased the maximum force resistance by approximately 14–20% compared to native vaginal tissue alone. Specifically, the cross-shaped mesh significantly outperformed the quadratic mesh, providing enhanced mechanical reinforcement by up to 30%. Numerical simulations closely matched the experimental results, with discrepancies of around 6% for vaginal tissue, 7% for mesh implants alone, and approximately 14% for mesh-reinforced tissue. These computational models effectively predicted mechanical behavior, indicating their potential in reducing the reliance on extensive in vivo testing.
The study concludes that biodegradable PCL meshes present a highly promising alternative to conventional synthetic meshes for POP treatment. Optimized mesh geometries significantly enhance mechanical support, potentially mitigating complications associated with current synthetic implants. Additionally, validated computational models offer a reliable framework for preclinical assessments, decreasing the dependence on animal testing and accelerating the development of future urogynecological implants. Future research should prioritize in vivo studies to evaluate long-term biocompatibility, degradation profiles, and clinical outcomes.

Pelvic organ prolapse (POP) is one of the most common pelvic floor dysfunctions (PFD), described as the protrusion of the pelvic organs through the vaginal walls. Recently, cases have been rising, affecting women's quality of life. Severe cases often require surgical mesh implants, which can cause complications like tissue erosion and infection, leading the FDA to ban transvaginal meshes for POP [1,2]. It is believed that the reported complications are most likely due to insufficient biocompatibility and inappropriate biomechanical properties of these implants [3]. An ideal mesh should be biocompatible, nontoxic, and have antibacterial properties and appropriate mechanical properties. To address this, polycaprolactone (PCL) mesh implants were produced via melt electrowriting (MEW), enabling the printing of complex 3D structures with high resolution at a low cost from melted polymers [4]. After, the meshes were evaluated mechanically and coated with azithromycin, an antibiotic for genitourinary infections.

Uniaxial tensile tests were performed on the meshes, allowing to obtain average stress-strain curves. Then the meshes went through cyclic tests to assess the long-term behavior of the meshes over 100 cycles. Results showed that while all PCL meshes had similar behaviour, those with 1 mm pores sustained higher stress, had a higher resistance to plastic deformation and could endure 19% more stress than the 1.5 mm pore size ones. However, the 1.5 mm pore-size meshes had mechanical properties closer to vaginal tissue but still remain stiffer. Regarding the cyclic tests, the material showed cyclic creep behavior, since the plastic deformation accumulates, and the curves do not go back to the same displacement [5]. It revealed initial damage, with the first cycle causing more damage to the mesh and hardening during plastic deformation. The tensile tests performed after cyclic deformation confirm increased stiffness, as Young’s Modulus rose between 19.2% and 29.3%.

Zone inhibition and biofilm assays evaluated azithromycin’s effectiveness against bacterial infection. Even though FTIR analysis could not confirm antibiotic incorporation, the drug-coated meshes show inhibitory activity against Escherichia coli biofilm formation and Methicillin-susceptible Staphylococcus aureus (MSSA) in its planktonic state. However, it revealed no antimicrobial effect against Methicillin-resistant Staphylococcus aureus (MRSA), which was expected, since MRSA is known as one of the most serious antibiotic-resistant bacteria. Scanning electron microscopy supported these findings. Overall, the results obtained prove that azithromycin incorporation in the PCL meshes was successful, suggesting that MEW-fabricated PCL meshes coated with azithromycin hold promise as improved implants for POP treatment.

This way, drug incorporation allied with meshes with appropriate mechanical characteristics has the potential to obtain a better outcome in surgical mesh implantation for treating POP.

[1] E. Mancuso, C. Downey, E. Doxford‐Hook, M.G. Bryant, P. Culmer, J Biomed Mater Res B Appl Biomater 108 (2020) 771–789.

[2] Food and Drug Administration, (n.d.).

[3] R. Rynkevic, M.E.T. Silva, P. Martins, T. Mascarenhas, J.L. Alves, A.A. Fernandes, Mater Today Commun 32 (2022).

[4] P.D. Dalton, Curr Opin Biomed Eng 2 (2017) 49–57.

[5] A. Głuchowski, W. Sas, Materials 13 (2020) 3907.

The mechanical response of human spinal dura mater under higher strain rates remains incompletely characterized, despite the critical protective role of this fibrous membrane. In the present investigation, samples were harvested from six human donors with a mean age of 79.33 years and prepared in a longitudinal orientation along the spinal axis. Dog-bone-shaped specimens were subjected to uniaxial tensile loading at nominal strain rates of 0.5, 10, and 25 s⁻¹ using a custom-built testing apparatus. Parameters such as elastic modulus, ultimate tensile strength (UTS), and stretch at failure were evaluated potential rate-related effects. Morphometric measurements indicated that posterior specimens exhibited statistically greater thickness and cross-sectional area than anterior ones. However, no definitive increase in mechanical strength was observed on that basis alone.

Strain-rate-dependent stiffening of the dura mater was recorded, with statistically higher elastic modulus values measured at 10 and 25 s⁻¹ relative to 0.5 s⁻¹. At 25 s⁻¹, the mean elastic modulus exceeded 265 MPa, compared to approximately 172 MPa at 0.5 s⁻¹. Ultimate tensile stress values showed overlapping standard deviations but trended toward higher mean levels at 10 s⁻¹. Furthermore, failure stretch appeared slightly lower at 25 s⁻¹, suggesting a reduced ability to sustain elongation under rapid loading conditions. These observations align with previously noted viscoelastic phenomena in other collagenous tissues and underscore the importance of testing spinal dura mater at physiologically and clinically relevant strain rates.

A visco-hyperelastic constitutive framework was used to capture the nonlinear and time-dependent properties of the dura. This approach incorporated transversely isotropic material behavior, reflecting the predominant collagen fiber orientation in the craniocaudal direction, along with rate-dependent viscous terms. Excellent correlations were achieved between the proposed material model and the experimental stress–stretch data across the tested strain rates, as confirmed by high coefficients of determination (adjusted R² > 0.98) and low error metrics.

These results expand the existing knowledge base by characterizing human spinal dura mater at strain rates relevant to high-speed trauma. Inclusion of the derived constitutive model in finite element analyses is anticipated to enhance the biofidelity of human body models, particularly in automotive crashworthiness research or other high-impact applications. In addition, the demonstrated regional geometric differences suggest that more targeted investigations, including biaxial experiments and correlation with tissue microstructure, may further elucidate localized mechanical variations. Overall, the findings highlight the pronounced rate-dependent stiffening of spinal dura mater and contribute essential data for improved simulation of dynamic spinal loading scenarios. Such observations underscore the clinical importance of understanding how rapid loading events may compromise the dural barrier. Further integration of microstructural analyses, including collagen fiber mapping, may clarify the relationship between regional tissue composition and local mechanical outcomes.

Pelvic organ prolapse (POP) is a prevalent condition that significantly impacts the quality of life for women. It occurs when one or more pelvic organs protrude through the vagina, outside the pelvis. Research indicates that up to 50% of women may experience POP at some point in their lives. The risk factors include vaginal childbirth, advancing age, obesity, or increased intra-abdominal pressure [1].

Apical prolapse, particularly affecting the uterus or vaginal vault, is a common form of POP often associated with weakened support structures at the apex, such as the uterosacral ligaments (USLs) and cardinal ligaments (CLs). These ligaments provide structural support, helping maintain the position of distensible organs such as the bladder, vagina, rectum, and uterus.

The number of pelvic floor dysfunctions, including pelvic organ prolapse, is expected to rise, increasing the demand for effective treatments. While synthetic meshes have been widely used, some were banned from the market by the FDA due to safety concerns [2]. Therefore, new tools to improve our understanding of this issue are crucial.

Recent studies on 3D printing have focused on developing biodegradable meshes using Polycaprolactone (PCL), a biocompatible and FDA-approved polyester. PCL degrades within 2–3 years and may hold promise for POP repair surgery, potentially reducing mesh-related complications (MRCs) such as chronic pain.

Proper anchoring and secure suture fixation are essential for preventing post-surgical MRCs. These techniques ensure mesh stability until full tissue integration is achieved. Mesh anchoring failure has been reported in 38% of patients, with an average occurrence of 1.8 years after implantation. Therefore, a comprehensive understanding of the implantation process is crucial for optimizing surgical outcomes.

In the present study, computational models of biodegradable porous mesh implants with square and sinusoidal geometries were developed to mimic the function of the USLs and CLs, which are critical for supporting pelvic organs. These computational implants were integrated into a computational pelvic cavity model to evaluate the superior-inferior displacement of the anterior vaginal wall during the Valsalva maneuver (VM) at different levels of ligaments impairment: 50%, 90%, and 100% (total rupture).

The results showed a baseline supero-inferior displacement of 6.38 mm in the healthy model. For the USLs, a 50% impairment increased displacement to 7.03 mm while a total rupture resulted in a displacement of 10.77 mm. For the CLs, impairment of 50% led to displacements of 6.74 mm while total rupture resulted in a displacement of 8.58 mm.

The integration of computational implant models reduced the anterior vaginal wall displacement across all impairment scenarios. Implants USLs-mimicking with square geometry, restoring up to 9% of displacement in cases of total USLs rupture compared to the asymptomatic model. When impairment was applied simultaneously to both ligament types (CLs and USLs), it was observed that the USLs primarily provide support, whereas the CLs mainly contribute to stabilization.

These findings underscore the promising potential of biodegradable meshes in improving POP repair.

Keywords: pelvic organ prolapse, apical ligaments, biodegradable implants, computational models

[1] Carroll et al.,PloS One,17:e0276788,2022.

[2] Food and Drug Administration,FDA actions on surgical mesh,2019.

1. Introduction

Eye movements are essential for visual integration, from fixational eye movements to saccades. Several studies have shown dependence between eye movements and ocular tissues strain (Demer, 2016). Integration of large eye movements in Finite Element Model (FEM) was previously addressed on simplified geometries (Karimi et al., 2017). Predicting activations parameters within a continuum extra ocular muscle model is a challenge since in-vivo measures are difficult. We propose a FEM that includes sclera, optic nerve and horizontal rectus muscles to implement new range of movements.

2. Methods

We retrieved an Abaqus assembly from a study on the ocular adduction by active extraocular muscle (EOM) contraction (Jafari et al., 2021). We assumed all tissues except the EOMs to be homogeneous, isotropic, incompressible, and hyperelastic based on previous modelisation (Jafari et al., 2021). A user subroutine models muscle contraction via EOM material properties. The quasi-incompressible strain energy density formulation captures both active and passive muscle behaviors (Grasa & Calvo, 2021). Muscle fibers anisotropy is accounted with an active stretch along the muscle fibers.

3. Results and discussion

Based on previous methodologies (Karimi et al., 2017), the model will be validated to reproduce horizontal saccadic movements where tensions in extra-ocular muscles have been quantified. Activations of the rectus muscles for saccadic movements exceeding previous range of motion are established. The principal strains will be quantified for saccadic movements in the peripapillary sclera.

4. Conclusions

Simulations of movements and corresponding activations will be extended through inverse simulations based on movements captured with an eye tracker. Positions implying important strains will be assessed, based on previous quantification of strains in eyes (Demer, 2016; Jafari et al, 2024).

Acknowledgements:

We thank S. Jafari and J. Demer for the Abaqus assembly and the 3D mesh.

Funding:

This project was partially funded by the French National Agency of Research and Technology via the convention Cifre 2023/0280. Several authors are employees of EssilorLuxottica.

References:

Demer, J. L. (2016). Optic Nerve Sheath as a Novel Mechanical Load on the Globe in Ocular Duction. Investigative Opthalmology & Visual Science, 57(4), 1826. doi: 10.1167/iovs.15-18718

Karami, A., Eghtesad, M., & Haghpanah, S. A. (2017). Prediction of muscle activation for an eye movement with finite element modeling. Computers in Biology and Medicine, 89, 368‑378. doi: 10.1016/j.compbiomed.2017.08.018

Jafari, S., Lu, Y., Park, J., & Demer, J. L. (2021). Finite Element Model of Ocular Adduction by Active Extraocular Muscle Contraction. Investigative Opthalmology & Visual Science, 62(1), 1. doi: 10.1167/iovs.62.1.1

Grasa, J., & Calvo, B. (2021). Simulating Extraocular Muscle Dynamics. A Comparison between Dynamic Implicit and Explicit Finite Element Methods. Mathematics, 9(9), 1024. doi: 10.3390/math9091024

We present a multiphysics computational framework for simulating the fully coupled interplay of cardiac electrophysiology, active and passive muscular mechanics, blood dynamics and valve kinematics in the human heart [Bucelli et al., IJNMBE, 2023]. The framework supports relevant sources of feedback between the models, including electro-mechanical and mechano-electrical coupling and fluid-structure interaction (FSI). We account for the geometric coupling between fluid and solid through the body-fitted ALE approach. The fluid domain is displaced with a non-linear mesh motion operator, which provides the robustness needed to deal with the large displacements characterizing the heartbeat. We account for valves through the resistive immersed implicit surface (RIIS) method [Fedele et al., BMMB, 2017], adapted to account for the FSI model of cardiac walls and blood flow. The heart model is coupled at inlets and outlets to a closed-loop lumped parameter model of the circulatory system [Fedele et al., CMAME, 2023].

We solve the coupled problem through the finite element method, treating the multiphysics coupled system through a segregated-staggered time discretization, wherein different submodels are discretized with different temporal resolutions to account for their particular accuracy requirements. The FSI subproblem is addressed with a monolithic solver that balances computational efficiency, accuracy and robustness.

The resulting computational framework is highly modular, meaning that its complexity may be adjusted depending on the need of each application. We demonstrate this through applications to cardiac electro-fluid-structure interaction of a realistic left heart, whole-heart cardiac electromechanics, and standalone hemodynamics models. The latter can be driven by cardiac electromechanics (resulting in a one-way coupled FSI system) or by a patient-specific calibration of the circulatory system. This is applied to proof-of-concept simulations of atrial hemodynamics under fibrillation coupled to models of thrombus formation.

The EuroHPC JU project dealii-X grant no. 101172493, funded under the HORIZON-EUROHPC-JU-2023-COE-03-01 initiative, is acknowledged.

The Woven EndoBridge (WEB) is an endovascular device increasingly used for treating intracranial aneurysms. However, selecting the appropriate WEB size preoperatively remains challenging, as it deforms significantly upon deployment depending on the aneurysm morphology. Approximately 20% of cases require redeployment due to size mismatch. Furthermore, while hemodynamic alterations are considered key to aneurysm occlusion, the number of studies quantifying changes before and after WEB deployment is limited. This study aimed to predict the post-deployment geometry of the WEB using preoperative imaging and to evaluate hemodynamic changes induced by the device. A patient-specific model of unruptured intracranial aneurysm that achieved complete occlusion after WEB deployment was analyzed. The aneurysm and arteries were reconstructed to generate a computational grid. For the post-deployment analysis, a virtual WEB geometry was modeled using an original Virtual WEB Deployment Simulation technique, which accounts for the deformation of the WEB based on its geometrical configuration and preoperative intracranial aneurysm morphology. Computational fluid dynamics (CFD) analysis were conducted for both before and after WEB deployment. The flow field was assumed to be incompressible and laminar, with blood modeled as a Newtonian fluid having a density of 1100 kg/m³ and a viscosity of 0.0036 Pa·s. The arterial wall was assumed to be rigid. As boundary conditions, the average diastolic mass flow in a healthy adult’s internal carotid artery was used at the inlet, and fixed static pressure of 0 Pa was applied at the outlet. We evaluated the mean blood flow velocity within the aneurysm, mean pressure, and mean wall shear stress (WSS) on the aneurysm wall before and after WEB deployment. Additionally, the rate of change for these parameters were calculated. As a result, the simulation qualitatively reproduced the geometrical characteristics of the deployed WEB, with the wires expanded along the aneurysmal wall in a manner closely resembling clinical deployment. However, geometrical discrepancies were observed near the neck region, indicating the need for further refinement of the modeling approach. Furthermore, after the WEB deployment, the mean blood flow velocity within the aneurysm decreased from 0.342 m/s to 0.159 m/s (−53.5%), and the mean WSS decreased from 7.99 Pa to 2.76 Pa (−65.5%). A reduction in pressure concentration at the aneurysm dome was also observed. These findings suggest that our Virtual WEB Deployment Simulation may enable accurate preoperative prediction of post-deployment WEB geometry and assist in selecting optimal size before surgery. Furthermore, a reduction in mean blood velocity within the aneurysm, the disappearance of high-pressure regions, and a decrease in mean WSS were observed after WEB deployment. These findings suggest that the reduction in blood flow velocity induced by WEB deployment may contribute to aneurysm occlusion.

The complex motility patterns of the stomach, characterized by antral contractions and tonic contractions, play crucial roles in mixing, grinding, and propelling food through the gastrointestinal tract. These mechanical processes, combined with chemical digestion, break down ingested food into smaller particles, increasing surface area for enzymatic action and nutrient absorption. Computational Fluid Dynamics (CFD) has emerged as a powerful in silico tool to complement in vitro and in vivo studies, offering detailed insights into gastric flow patterns and mixing mechanisms that are difficult to measure experimentally. CFD models can simulate the fluid dynamics of gastric contents, track chemical species, and analyze the impact of varying physiological parameters on digestion efficiency.

Existing CFD models of gastric motility have primarily represented gastric contents as Newtonian fluids, with food often modelled as a chemical species rather than explicit solid particles. This simplification limits the ability to study the mechanical breakdown of solid foods, a critical aspect of gastric digestion. Furthermore, many models represent the opening and closing mechanisms of the pyloric sphincter, instead assuming a constantly open or closed pylorus. Additionally, most studies simulate only brief periods of gastric activity, typically minutes rather than the hours required for complete digestion, mainly due to the computational costs. These limitations highlight the need for more comprehensive models that capture the full complexity of digestion.

To address these limitations, we have developed a coupled particle-based model using Smoothed Particle Hydrodynamics (SPH) and Lattice Spring Model (LSM) techniques. This approach is better suited to simulate the large amplitude contractions of the stomach wall and pyloric sphincter, as well as the mechanical breakdown of solid foods. Our SPH model of gastric emptying was implemented using the open-source LAMMPS package on an idealized stomach and duodenum geometry. The model incorporates boundary deformation patterns derived from a previously developed model of gastric motility and pylorus sphincter activity, allowing for detailed simulation of gastric motility.

The SPH model successfully captured key gastric flow patterns, including retropulsive jets and recirculation regions, which are essential for effective mixing and emptying. The simulated half-emptying time for water (13 minutes) aligns closely with in-vivo human studies. The ability of the CFD model to represent pyloric sphincter dynamics provides a more accurate representation of gastric emptying regulation. The solid-phase of digesta is represented using LSM to explicitly model food particle breakdown.

Future work includes developing subject-specific models using MRI data from an ongoing study investigating the gastric digestion of halloumi cheese. This study simultaneously measures motility patterns and digestion kinetics, providing valuable validation data for our SPH-LSM coupled model. Our approach offers a significant advancement in computational modelling of gastric function, bridging the gap between micro-scale processes and whole-organ mechanics, and providing new insights into smooth muscle organ dynamics.

Accurate assessment of rupture risk is critical for the clinical management of intracranial aneurysms, as rupture can result in life-threatening subarachnoid hemorrhage with a high mortality rate. Although unruptured aneurysms are increasingly detected due to advances in diagnostic imaging, low annual rupture rate poses a clinical dilemma regarding surgical intervention. Previous studies have reported that thin-walled regions (TWRs)―areas of reduced wall thickness and structural fragility―are likely sites for rupture initiation. However, non-invasive identification of TWRs remains challenging, since conventional imaging modalities cannot directly evaluate aneurysm wall thickness. Recently, computational fluid dynamics (CFD) analysis has been employed to investigate hemodynamic factors associated with TWRs. This study aimed to construct a predictive model for TWRs using CFD analysis and to assess its practical applicability through validation.

A total of 124 aneurysm cases treated with surgical clipping at two institutions were analyzed. At Institution A, 109 cases before May 2021 were used to build the predictive model, and 13 subsequent cases were used for validation. Two additional cases from Institution B were used for external validation. Patient-specific three-dimensional arterial geometries were reconstructed from angiographic imaging. CFD simulations were performed assuming incompressible and laminar flow modeled as a Newtonian fluid. A rigid wall was assumed, and the no-slip boundary condition was applied to it. The inlet boundary condition reflected typical pulsatile blood flow in healthy adults, while the outlet pressure was set to 0 Pa. Hemodynamic parameters such as Pressure Difference (PD*) and Wall Shear Stress Divergence (WSSD*) were calculated on the aneurysm wall. Intraoperative microscopic images were used to quantify wall color via the comprehensive Red (cR) value, with the top 25% of pixels defined as TWRs. A prediction model for TWRs was formulated using logistic regression based on hemodynamic parameters. The Risk of TWR (RoT) was calculated for the aneurysm surface, and the top 25% RoT regions were defined as predicted TWRs. Prediction accuracy was evaluated using Inclusion Rate, which measures the extent to which the predicted region overlaps with the actual TWRs.
Logistic regression showed that PD* and WSSD* were significantly higher in TWRs than in thick-walled regions. A prediction formula for TWRs was constructed based on multivariate analysis. Application of this model to the training dataset yielded an average Inclusion Rate of 77.2%, while validation on cases from Institutions A and B resulted in rates of 77.9% and 75.2%, respectively. These results indicate comparable predictive performance across datasets. In high-Inclusion Rate cases, areas with high PD* and WSSD* qualitatively matched intraoperatively observed TWRs, reflecting localized stress concentrations in both perpendicular and tangential directions. In low-Inclusion Rate cases, often involving large aneurysms with stagnant flow, the prediction parameters did not exhibit high values, which likely made it difficult to identify TWRs. Incorporating morphological factors of intracranial aneurysms may improve the accuracy of TWR prediction.

In conclusion, this study demonstrates the feasibility of CFD-based prediction of TWRs in intracranial aneurysms across multiple datasets. The proposed method may support more accurate preoperative risk assessment and help guide treatment decisions to improve patient outcomes.

Introduction

In orthopaedics, traditional treatments generally involve the use of techniques and medical devices developed on the basis of population samples. However, they are unfortunately unlikely to account for the actual multiple characteristics specific to each individual patient, including all morphological and functional parameters of the anatomical complex under treatment. As surgical instrumentation and implantable devices are generally not sized to the patient's specific anatomy, which can lead to inaccurate implant positioning, impaired postoperative functional performance, and potentially to the surgical failure. The extraordinary advances in medical-imaging techniques and the advent of 3D printing, using basic biocompatible materials, i.e. polymeric/metal filament or powders, makes it possible to manufacture unique, highly complex medical devices. Hence, it is now possible and accessible to design and manufacture patient-specific surgical instrumentation and implant components that are well adapted to the patient's actual morphology. The aim of this study is to provide an in-depth, orthopaedic-focused view of the real potential and applications of 3D printing in the field of personalized surgery.

Advancements and Applications

Innovative solutions for the definition of fully customized medical devices for clinical applications have recently been developed at the Rizzoli Orthopaedic Institute in Bologna, Italy. These can be applied in conservative treatments or surgical procedures, these including minimal and massive skeletal reconstructions. A number of applications have been conducted in preclinical investigations and early clinical cases. Typically, they are based on reliable and accurate anatomical modelling using state-of-the-art medical imaging systems for morphological reconstruction. More specifically, these now include innovative weight-bearing CT based on cone-beam technology and EOS imaging, both with reduced radiation dose compared to standard devices and therefore safer for the patients and in paediatric applications, as well as dual-energy CT and 3.0T MRI supported by machine-learning protocols.

Preclinical applications included in-vitro testing of novel joint prostheses prototypes, investigation of lattice structures for bone growth into implant using cell cultures, sensitivity analysis on anatomical reconstruction from medical-image segmentation, and identification of the most appropriate clinical-oriented 3D printing parameters. The most recent clinical experiences include customized total knee and ankle replacement, personalised femur and tibia osteotomy for correction of varus-valgus deformity of the knee and ankle, massive pelvis reconstruction in oncologic patients, of spine deformity correction, vertebrectomy with subsequent reconstruction, patient-specific tibial intercalary segmental reconstruction, coronoid fracture repair, and many others. In all these applications, starting from CT/MRI-based morphological reconstructions of the patient's anatomy under treatment, the customised procedures involve surgical planning, design and EBM/SLM-based 3D printing, using biocompatible metal alloy powders, of the implantable devices and of the necessary surgical instrumentation. Other experiences have led to 3D printing of bio-models, which have proven to be essential in medical training, for preoperative planning and device preparation. The basic experiences principles can ideally be transferred to other medical branches.

Concluding Remarks

3D printing now seems clearly essential for the customization of orthopaedic treatments. We are confident that this technique will optimize patient-to-doctor communication, final patient performance, and medical training, resulting in cost-effective solutions for health care systems and industries.

Bone fractures are a common and a major cause of disability and death worldwide, and up to 10% of all fractures fail to heal normally (called “fracture non-unions”). Ultrasound imaging is safe, cost-effective, compatible with metal implants, and widely used to diagnose soft tissue injuries. Importantly, it holds significant potential to revolutionize fracture care by enabling early and routine monitoring and assessment of fracture healing, which for patients at high-risk of non-union, could mean earlier detection and clinical intervention of poor healing fractures that would prevent advanced clinical state and prolonged patient suffering, and reduce extensive waiting times for treatment, which would reduce costs for the healthcare system and patients. However, as ultrasound imaging is not routinely used in orthopedic clinics, barriers include difficulty acquiring and interpreting ultrasound images, lack of standardized scanning guidance and unclear terminology for bone and clinically important features of fracture healing. Previous efforts to use ultrasound imaging for musculoskeletal applications have demonstrated large intra‐examiner and inter‐examiner reliability variance for semi-quantitative and continuous measures, and image processing was slow as segmentation is often manual (or semi-automated). Therefore, there exists strong need to develop automated approaches to reduce the observed variance and to automate the segmentation procedure. We have established terminology and reporting guidelines for bone healing on ultrasound imaging; including expert agreement on key bone healing features from orthopaedic surgeons, radiographers and medical physics experts, and we have developed a Python-based, multi-label classification machine learning (ML) algorithm using these consensus labels to aid clinical interpretation of ultrasound scans. We have assess 2D US image label classification performance of four pre-trained Convolutional Neural Networks (CNNs) that have previously demonstrated good bone segmentation ability within US imaging in recent studies (VGG-16, VGG-19, Resnet50 and deeplabv3) using a dataset consisting of 734 2D ultrasound images of two superficial and frequently fractured long bones, the tibia (3 x longitudinal scan sweeps, 366 2D images) and the clavicle (3 x longitudinal scan sweeps, 368 2D images) acquired using an ultrasound scanner (L15 HD3, Clarius, Vancouver, BC, Canada). While all four tested algorithms performed with relatively high precision, recall/sensitivity and F1 scores, the deeplabv3 algorithm demonstrated the best segmentation performance across all metric parameters for bone segmentation on 2D US imaging, followed closely by Resnet50, then VGG-19 and VGG-16. Augmenting the dataset resulted in notable improvements to mean class accuracy particularly for VGG-16 and deeplabv3 and marginally decreased performance for VGG-19 and Resnet50. Overall, dataset augmentation resulted in marginal overall improvements. This is the first study to compare the performance of four different and state-of-the-art clinically used CNNs to identify the algorithm demonstrating the best segmentation performance for bone on US imaging. The results presented are significant for inter-disciplinary biomedical engineers to design efficient, accurate and robust automated ML-based segmentation models for clinical orthopedic applications.

Muscle strength assessment is a fundamental outcome measure for evaluating shoulder function in patients with musculoskeletal disorders. Hand-held dynamometers (HHDs) are widely used in clinical practice due to their ease of use, cost-effectiveness, and portability. However, factors such as examiner strength, subject positioning, and HHD placement often compromise their reliability. To address these limitations, recent advancements in orthopedic bioengineering have explored novel measurement devices, including load cells, which provide higher accuracy and improved data consistency.

This study proposed an innovative system that integrates an HHD with a load cell mounted on a rigid, 3D-printed, modular support, designed to reduce operator dependency and enhance measurement precision. By standardizing force measurements and eliminating variability due to manual positioning and tester influence, this configuration aims to improve the reproducibility of muscle strength assessments. Furthermore, the system enables experimental trials in which the load cell and HHD function in a serial configuration, allowing for a direct comparison of their precision and reliability in force measurement.

The system was validated through static weight tests ranging from 9.81 to 98.10 N, based on previous experiments conducted with healthy volunteers and literature studies involving patients with musculoskeletal shoulder disorders. Validation tests using static weights demonstrated good consistency between the two devices, though discrepancies became more pronounced as the weight increased. A feasibility assessment was also conducted with healthy volunteers, who underwent standardized strength evaluations under controlled conditions to assess the system's reliability and validity. Results confirmed the system's capability to measure shoulder muscle strength and facilitate comparisons across different measurement techniques. Specifically, the Bland-Altman analysis indicated a small systematic bias (mean difference: 1.6 N) and narrow Limits of Agreement (LOA: -3.8 N to 7.0 N). The Mean Absolute Error (MAE: 2.9 N) and Mean Absolute Percentage Error (MAPE: 9.4%) further demonstrated acceptable error levels in both dominant and non-dominant arms.

The integration of advanced bioengineering solutions into muscle strength assessment represents a significant step toward enhancing the accuracy, reliability, and standardization of shoulder function evaluation. By minimizing operator dependency and optimizing measurement protocols, this approach offers a robust framework for clinical assessments, rehabilitation monitoring, and postoperative outcome evaluations. Additionally, it offers a more objective and data-driven approach for evaluating musculoskeletal impairments, facilitating early diagnosis and personalized treatment planning. Future large-scale studies are needed to validate the system across diverse populations and clinical environments, ensuring its effectiveness in real-world applications. The proposed system demonstrates strong potential for clinical adoption, particularly in the continuous monitoring and diagnosis of musculoskeletal shoulder disorders, where precise and reproducible strength measurement is critical for guiding therapeutic interventions and optimizing patient outcomes.

Acknowledgement: Funded by the European Union - Next Generation EU - NRRP M6C2 - Investment 2.1 Enhancement and strengthening of biomedical research in the NHS (Project no. PNRR-MAD-2022-12376080 - CUP: F83C22002450001).

Knee osteoarthritis (KOA) is a common degenerative musculoskeletal condition among the elderly, often resulting in chronic pain and reduced mobility. Numerous studies have shown that KOA significantly alters knee biomechanics, leading to changes in movement patterns. Therefore, analyzing lower limb biomechanics is critical for identifying parameters that describe functional limitations and quantifying the effectiveness of therapeutic interventions. To date, accurately characterizing the complex kinematics of the knee joint in patients with KOA remains a challenge. Optoelectronic motion capture systems, currently considered the gold standard, can provide quantitative assessments of 3D kinematics. In particular, gait analysis plays a crucial role in providing a detailed understanding of changes in movement and joint mechanics in subjects with KOA.

Over the past decades, a wide range of biomechanical models have been developed to estimate joint kinematics. Several comparative studies have been conducted on healthy populations to assess the performance of different gait models and marker configurations. These investigations have yielded valuable insights into the reliability and repeatability of kinematic outputs. However, limited attention has been given to evaluating such models in patients with KOA, who often present with knee deformities and soft tissue swelling, which can affect marker positioning and joint center estimation. Consequently, there remains a need to investigate the agreement between different models, specifically in this population, where such anatomical variations can increase uncertainty in kinematic results and impact clinical interpretation.

This study aim to investigate the inter-model agreement between two widely implemented gait models: the anatomically-based Istituto Ortopedico Rizzoli (IOR) model and the cluster-based Calibrated Anatomical System Technique (CAST) model, using both six degrees of freedom (6DoF) and inverse kinematics (IK) computational approaches.

A total of 16 patients was enrolled in this study. Each patient completed six walking trials on a 3-meter walkway at their own pace, with both the IOR and CAST marker sets applied simultaneously. Agreement between models was then quantitatively assessed using intraclass correlation coefficients (ICCs) and mean absolute differences (MD).

Results indicated high consistency across all models in measuring sagittal plane knee dynamics, with ICC values consistently exceeding 0.75 and mean absolute differences below 2.4°, demonstrating reliability during all phases of gait (heel strike, toe-off, loading response, terminal stance, and maximum swing). However, considerable variability emerged in frontal and transverse plane movements. Notably, frontal plane analyses revealed mean absolute differences as large as 12.7° during the maximum swing phase between 6DoF configurations of CAST and IOR models. Comparatively smaller discrepancies (≤5.8°) were noted between IK configurations.

The variations observed between IOR and CAST models, as well as between the 6DoF and IK computational methods in the frontal and transverse planes, highlighted the importance of refinement and validation of gait analysis methodologies. Careful model selection is crucial in clinical and research settings, particularly for populations with complex and variable joint dynamics like individuals with KOA.

Acknowledgment: Funded by the European Union - Next Generation EU-NRRP M6C2-Investment 2.1 Enhancement and strengthening of biomedical research in the NHS (Project no. PNRR-MCNT2-2023-12378237-CUP: F87G24000130006)

Lungs are lined with a thin layer of mucus which aids in clearing away foreign particulate. To remove this dirty mucus from the lung, bronchi are lined with cilia: small hair-like structures which beat, forcing the mucus to flow upwards. Due to the difficulties associated with direct observation of mucus flow, in-vitro methods are necessary to investigate mucociliary clearance and how it is affected by cilia behaviour and mucus composition. One approach to studying cilia has been the use of artificial microfabricated cilia, which actuate under a varying magnetic field¹.

Our cilia use soft lithography in order to fill moulds with iron microparticles, as well as polydimethylsiloxane, which is then cured². By creating a rotating magnetic field using permanent magnets mounted on a motorized axis, we are able to asymmetrically actuate our cilia. The cilia motion comprises a whipping active stroke, which is capable of pushing fluid forward, coupled with a recovery stroke, where the cilium slowly swings around to its original position. We measured the pumping capacity of these actuated cilia by particle tracking velocimetry in closed channels. The asymmetric movement of the cilia results in pulsatile flow within the channel³. Using micellar solutions of different concentrations, we show the diminishing pumping rate with increased fluid viscosity.

Our artificial cilia exhibit both orientational and temporal asymmetry. Notably, two separate cilia pathways are identified in the same tests. By varying the position of our magnetic cilia relative to the magnetic field, we are able to identify three distinct behaviours. Further from the magnets, we observe a symmetric pillar rotation. When closer to the magnet, the cilia motion develops either a simple, or compound buckling behaviour. Beam simulations which couple the magnetic forces and elastic forces acting on our artificial cilia, based on previous analysis of our artificial cilia³, are capable of reproducing the three distinct behaviours. This model highlights the important of the magneto-elastic number in artificial cilia systems.

Following this, we use a coupled fluid-structure simulation to investigate separately the importance of cilia pathway versus the velocity of our cilia. Comparing the flow caused by a single cilium with the Reynolds number of the cilium, we highlight the importance of temporal asymmetry in cilium pumping. This dependence upon the Reynolds number is one contributing factor which helps explain the diminishing pumping efficiency of more viscous solutions.

[1] Zhang S. et al. (2018) Sensors and Actuators B 263 614-624
[2] Bolteau B. et al. (2023) ACS Appl Mater Interfaces 15(29):35674-35683
[3] Moore et al. (2025) arXiv:2504.0064

The lungs are highly porous organs, and their mechanical response is critical for sustaining life. This response is significantly influenced by the surface tension resulting from the air-liquid interface in the alveoli. One of the most remarkable features of alveolar surface tension is its dynamic behavior, which can be explained by the activity of pulmonary surfactant, a complex compound that reduces surface tension, preventing alveolar collapse and facilitating normal breathing. Despite its crucial role, surfactant-dependent surface tension has typically been neglected in continuum models of the lungs, possibly due to its complex multiscale physicochemical nature.

In this research, we developed a continuum model for lung parenchyma that incorporates the surfactant-dependent surface tension in the alveoli, allowing us to predict the hysteretic response of the lungs. To this end, we adopted a poromechanical framework in which the parenchyma domain interacts with the air phase. To ensure thermodynamic consistency, we considered the Clausius-Duhem inequality for an isothermal porous medium. Using a standard Coleman-Noll procedure and taking surfactant mass as an internal variable, we derive an expression for the stress tensor that includes a collapsing term related to the surfactant-dependent surface tension. The resulting governing equations were solved using a non-linear finite element scheme on patient-informed lung geometries [1].

To test our model, we simulate the supersyringe procedure on an excised lung, tracking the temporal evolution of volume, airway pressure, and surface tension. From these time series, we constructed the quasi-static pressure-volume (P-V) curve. Different pressures were needed for inflation and deflation of the P-V curve, resulting in markedly hysteretic behavior. Changes in the slope of P-V curves, also known as lung compliance, were observed. We also derived the quasi-static surface tension-volume curve, which exhibited highly non-linear behavior that seems to be related to the shape of the P-V curve.

Our model captures the effect of surfactant dynamics and the resulting alveolar surface tension on the tissue and organ response. Given the scale differences between surfactant and whole organ, surfactant activity was naturally integrated into a continuous framework using an internal variable approach. Remarkably, our simulations recovered the hysteresis of the pressure-volume curves, a well-known characteristic of human lungs typically associated with surfactant behavior. Furthermore, our results suggest that the shape of the surface tension-volume curve strongly influences the compliance of the P-V curve. These observations align with experimental setups of excised lungs and reflect the role of surfactants in lung mechanical response [2]. We envision that our framework will facilitate lung simulations where surfactant-related phenomena are directly incorporated, with essential applications for modeling respiratory diseases and lung responses to mechanical ventilation.

Acknowledgments: This work was funded by the National Agency for Research and Development (ANID) of Chile through the grant FONDECYT Regular #1220465. NA-R acknowledges the support of the graduate fellowship ANID BECAS/DOCTORADO NACIONAL 21212320.

References: [1]-Avilés-Rojas, N., & Hurtado, D. E. (2022). Whole-lung finite-element models for mechanical ventilation and respiratory research applications. Front Phys, 13, 984286.[2]-Harris, R.S., 2005. Pressure-volume curves of the respiratory system. Respiratory care 50, 78–99.

Objective: Three dimensional magnetic resonance spirometry enables ventilation to be characterized on a regional scale using dynamic lung MRI acquisitions [1]. We are currently developing a biomechanical model of the ventilation to provide basic understanding of the underlying mechanical processes and to generate a patient-specific digital twin for personalised monitoring of respiratory pathologies. In this work, a simulation of airflow in the human bronchial tree is carried in one dimension using the incompressible Navier Stokes equations.

Methods: A one dimensional model is a first approach to address air flows in the proximal airways as it represents a good compromise between accuracy and numerical efficiency. Lung volumes, acquired with MRI, were segmented at the functional residual capacity down to the lobar level using nnUNet [2]. Gas flows were simulated using the SimVascular 1D solver translated to the gaseous phase [3] in bronchial trees that were generated within the geometrical boundaries of the segmented lung lobes [4]. The model assumes rigid airways. The model was tested using a simple case study. First simulations were performed while imposing a constant airflow of 250 cm3/s at the inlet of every lobar branch and applying a Windkessel model [5] to the outlets with 1193,355342 dyn.s/cm5 for the proximal resistance, 3,71x10-5 cm5/dyn for the capacitance and 14718.049219 dyn.s/cm5 for the distal resistance. These parameters do not correspond to those of the airways, and estimation of these parameters is ongoing, but a choice had to be made in order to launch our simulations. The pressure distribution was then computed throughout the bronchial tree.

Second, we make use of these pressures as boundary conditions to the model before running the simulations again to eventually infer the flow at the input of the model.

Conclusion: This work represents a first step towards a more complete and personalised model of pulmonary ventilation based on MRI lung volumes. Coupling with a poroelastic model of the lung parenchyma will enable us to recover the outlet pressures of the shafts and thus estimate the airflow in the airways like we did here over the test model [6].

References: [1] Boucneau T. et al., Sci Rep., 10, 9649, 2020 [2] Barrau N., 3D MR spirometry. PhD thesis, 2024. [3]Pfaller MR. et al., Ann Biomed Eng., 49, 3574, 2021 [4] Nousias S. et al., PLoS One, 15, 2020 [5] Vignon-Clementel IE. et al, Computer Methods in Biomecanics and Biomedical Engineering, 13, 625, 2010 [6] Patte C. et al. Biomech Model in Mechanobiol., 21, 527, 2022.

Acknowledgment : European Union Horizon Europe Research and Innovation Program with the agreement N° 101099934 (V|LFSpiro3D)

Introduction:
Biomechanical computational models should be largely patient-specific to provide deeper insight into the pathophysiology of abdominal aortic aneurysms (AAA) and to be relevant for future clinical diagnostics. The identification of individual material properties is crucial but remains challenging in vivo. Time-resolved 3D ultrasound combined with speckle tracking (4D-US) is a non-invasive imaging technique that provides full-field information of heterogeneous aortic wall strain distributions [1]. These strains are cyclic with respect to diastole. We present a novel 4D-US strain imaging approach for identifying parameters of a nonlinear, anisotropic material equation using only in vivo-accessible data, a further development of a previously presented method [2]. For this new method, we performed both in vitro and in vivo validations.

Material and Methods:
For in vitro validation, an intact porcine aorta was tested in an inflation-extension device. The vessel was immersed in 0.9% NaCl solution at 37°C and subjected to cyclic physiological pressure and length changes at 0.84 Hz. Forces, pressures, and deformations were recorded using two CMOS cameras and 4D-US imaging in parallel. A finite element analysis (FEA) model was generated from 4D-US images, with boundary conditions replicating the experimental setup. Parameters for the anisotropic Holzapfel-Gasser-Ogden (HGO) material equation were identified by (1) fitting to stress-stretch curves derived from mechanical data and (2) using the in vivo 4D-US strain imaging approach.

For in vivo validation, 4D-US data of two AAA patients (AAA1, AAA2) were acquired using a commercial 3D echocardiography system with a transthoracic probe. Diastolic and systolic blood pressures were recorded. During open AAA repair of the same patients, six wall specimen (three longitudinal, three circumferential) were harvested and tested in uniaxial tensile experiments, with deformations recorded optically. First, stress-stretch curves were calculated from mechanical data, and HGO parameters were identified by fitting the model to this mechanical data. Secondly, a patient-specific FEA model was created from in vivo 4D-US images, incorporating vessel wall and intraluminal thrombus. Physiological boundary conditions were applied, including measured diastolic and systolic pressures. HGO material parameters were identified using the in vivo 4D-US strain imaging approach.

Both in vitro and in vivo validations used an evolutionary self-adaptive Differential Evolution (JADE) algorithm for parameter optimization. The algorithm iteratively computed the material-dependent load-free geometry and then diastolic/systolic geometries using measured blood pressures, from which cyclic strains were derived. HGO parameters were identified by minimizing the error between measured (4D-US) and calculated (FEA) cyclic strains.

Results:
Comparison of stress-stretch curves calculated using the identified parameter sets of optical and 4D-US methods show coefficients of determination in the range R²=0.952−0.990. Performance was improved by a factor of seven compared to previous implementation.

For in vivo validation, AAA1 showed mean absolute percentage errors (MAPE) of 0.76%-0.07% when comparing stress-stretch curves calculated using the identified parameter sets of optical and 4D-US methods. AAA2 exhibited MAPE values between 0.51% and 17.83%. When averaging tensile test results from the four specimens, the stress-stretch curve aligned well with the 4D-US-derived global material behavior.

Conclusion:
We successfully validated a 4D-US strain imaging approach both in vitro and in vivo for identifying patient-specific parameters of an anisotropic material equation using only in vivo-accessible data. The method demonstrated high accuracy and feasibility for clinical application in AAA biomechanics.

Literature:
[1] Karatolios K, Wittek A et al., Ann. Thorac. Surg, 2013
[2] Wittek A et al. JMBBM 2016; 58: 122-138

Background: Abdominal aortic aneurysm (AAA) is a degenerative vascular disease characterized by significant changes in the geometry and microstructure of the aortic wall. AAA particularly affects elderly men, with a prevalence ranging from 1.3% to 8.9% in men over 60. Rupture of the aneurysm wall is associated with a high mortality rate of up to 80 %, so that adequate identification of patients at risk is crucial for their survival. Clinically, rupture risk is assessed using the maximum diameter criterion, which guides surgical decision-making. While statistically validated, this criterion lacks precision as it fails to account for patient-specific variations in aortic wall strength and local stress concentrations. The need for more patient-specific biomarkers is therefore widely acknowledged. It is hypothesized that Wall Motion Indices (WMI), derived from 4D ultrasound (4D-US)-based local strain measurements, can provide a more accurate assessment of patient-specific rupture risk by capturing aortic wall kinematics.

Methods: Preoperative data from 9 AAA patients were collected, including 4D-US strain imaging, CT angiography, maximum diameter, and blood pressure. Patients underwent open surgical repair, during which tissue samples were harvested for mechanical testing. WMI, describing the distribution of the aortic wall’s strain field, were computed from 4D-US strain data. The analysis focused specifically on wall regions corresponding to the locations from which tissue samples were obtained during surgery. These indices were correlated with a normalized experimental rupture potential (NERP), calculated as the ratio between in vivo wall tension (derived from Finite Element Analysis based on CT-A and mean arterial pressure) and the experimental failure tension (yield and maximum tension) determined via uniaxial tensile tests on the harvested tissue.

Results: Several WMI demonstrated significant correlations with NERP (r>0.75, p<0.05), particularly indices describing heterogeneous strain distributions within the aneurysmal wall. These WMI also correlated significantly with the experimental failure tension determined from uniaxial testing of the tissue samples (r<-0.67, p<0.05). Conversely, no significant correlation was found between maximum diameter and NERP, highlighting the limitation of size alone in predicting the individual mechanical failure risk.

Conclusion: This study indicates that 4D-US-derived WMI can provide patient-specific information correlated to rupture potential of AAA independent of the maximum diameter. The strong correlation between heterogeneous strain patterns and tension-based rupture potential highlights the potential of WMI to refine clinical decision-making. As a non-invasive modality, 4D-US derived WMI could potentially improve patient selection for elective repair, identifying individuals truly at high risk irrespective of diameter alone, thus personalizing AAA management. Further validation in larger patient cohorts is necessary to confirm these findings.

We propose a novel model for a vascularized poroelastic composite representing the myocardium which incorporates both mechanical deformations and electrical conductivity. Our structure comprises a poroelastic extracellular matrix with embedded vasculature and elastic inclusions (representing the myocytes) and we consider the electrical conductance between these two solid compartments. There is a distinct length scale separation between the scale where we can visibly see the connected fluid compartment (vessels) separated from the poroelastic matrix and the elastic myocyte and the overall size of the heart muscle.

We therefore apply the asymptotic homogenisation technique to derive the new model. The effective governing equations that we obtain describe the behaviour of the myocardium in terms of the zero-th order stresses, current densities, relative fluid-solid velocities, pressures, electric potentials and elastic displacements. It effectively accounts for the fluid filling in the pores of the poroelastic matrix, flow in the vessels, the transport of fluid between the vessels and the matrix, and the elastic deformation and electrical conductance between the poroelastic matrix and the myocyte.

This work paves the way towards a myocardium model that incorporates multiscale deformations and electrical conductivity whilst also considering the effects of the vascularisation and indeed the impact on mechanotransduction.

Background:

Gastric slow waves, generated and propagated by interstitial cells of Cajal (ICC), coordinate stomach motility and are essential for healthy digestion. Dysrhythmic slow wave activity is increasingly recognized as a biomarker for motility disorders such as gastroparesis and functional dyspepsia. Multiscale modeling of gastric electrophysiology linking structural, tissue, and organ-level dynamics enhances our understanding of gastric motility mechanisms and offers a pathway toward elucidating pathological changes and enabling non-invasive diagnostics.

Methods:

A combination of multiscale experimental and computational techniques was employed to investigate how ICC microstructure affects slow wave dynamics, and how these dynamics manifest at the organ and far-field levels. Detailed ICC networks in murine gastric tissues were imaged using multiphoton confocal microscopy. Region-specific continuum models of ICC networks in the antrum and corpus were constructed using machine learning-based segmentation and ICC density and morphological metrics were quantified. The impact of ICC structures on slow wave propagation was investigated using a self-excitatory ICC cell model. At the organ level, anatomically realistic stomach models were developed, and simulations of normal and dysrhythmic slow wave patterns informed by in-vivo high-resolution gastric mapping were performed. Additionally, forward simulations of biomagnetic fields were conducted using volume current source formulations. Simulated biomagnetic signals were compared against experimental recordings from porcine models undergoing simultaneous high-resolution serosal mapping and biomagnetic acquisition. Finally, an anatomically-constrained source imaging method was developed to map biomagnetic signals back to the stomach surface and reconstruct slow wave activation sequences.

Results:

Microscopic analysis revealed significant regional differences in ICC density and thickness across the stomach. The antrum exhibited the highest ICC density (13.0±5.1%), followed by the corpus (6.0±1.9%) and fundus (1.5±0.5%), which lacked myenteric plexus ICCs. These variations produced marked effects on slow wave speed and synchrony, with the antrum supporting faster, more coherent propagation. In the forward modeling analysis, simulated biomagnetic signals based on serosal activation patterns reproduced key features of experimental data, including waveform morphology and spatial distribution. Anatomically-based source imaging accurately reconstructed slow wave propagation patterns including simultaneous ectopic and normal wavefronts with mean localization errors within 6–14% of the longitudinal axis of the stomach.

Conclusion:

This multiscale framework, linking ICC structure to tissue-level slow wave behavior and far-field biomagnetic signatures, provides a foundation for non-invasive diagnosis of gastric dysrhythmias. The validated source imaging approach enables spatially resolved estimation of slow wave propagation from surface biomagnetic recordings offering a novel non-invasive clinical biomarker for the diagnosis of gastric motility disorders.

The stomach plays a fundamental role in digestion through accommodation, mechanical mixing, and chemical breakdown of food. Gastric peristalsis, a coordinated process of muscular contractions, ensures the efficient mixing and propulsion of food for subsequent digestion and absorption in the intestines. These contractions are governed by a complex electromechanical system, making a detailed understanding of gastric motility essential for diagnosing and treating gastrointestinal disorders such as gastroparesis, gastroesophageal reflux disease, and dyspepsia.

Despite its physiological importance, gastric biomechanics remains poorly understood, both in terms of biophysical mechanisms and computational modeling. While computational modeling has emerged as a promising, non-invasive tool for studying biomechanical systems, particularly when informed by medical imaging, there remains a scarcity of robust, physiologically accurate models that comprehensively capture the biomechanical complexity of gastric motility. Whole-stomach electromechanical models will provide valuable insights. However, whole-stomach models face critical computational challenges, including large deformations in nonlinear domains and patient-specific anatomical variability.

To address these challenges, we present a robust computational multiphysics framework for modeling human gastric motility using patient-specific geometries. Our approach integrates key biomechanical factors, including prestress, regional material properties, and non-uniform boundary conditions. Additionally, it incorporates an anisotropic active strain formulation within a homogenized constrained mixture framework, enabling more accurate simulations of structural and material heterogeneities in gastric biomechanics. Although patient-specific geometries can be reconstructed from medical imaging, crucial biomechanical properties—such as smooth muscle and collagen fiber orientations and the regional distribution of material parameters—are challenging to obtain directly. To overcome this, we introduce a novel workflow leveraging Laplace-Dirichlet-based reparameterization to infer physiological principal fiber orientations and spatial material distributions.

The proposed computational model successfully replicates essential features of gastric electromechanics, including slow wave entrainment and the physiologically accurate propagation of ring-shaped peristaltic contraction waves with large deformations. These findings emphasize the intricate coupling between electrical excitation and large-scale nonlinear tissue deformation. Moreover, the model’s adaptability to patient-specific geometries allows for precise representation of anatomical and physiological variations, which is crucial for understanding individual differences in gastric motility.

By integrating computational techniques with patient-specific imaging data, our framework enables large-scale in-silico studies of gastric motility. This approach provides a powerful tool for improving diagnostic accuracy, advancing the understanding of gastric motility disorders, and developing personalized treatment strategies. Ultimately, this work bridges the gap between computational modeling and clinical practice, offering a pathway toward more precise, patient-tailored interventions for disorders such as gastroparesis, gastroesophageal reflux disease, and dyspepsia.

Introduction: Contractions of the smooth muscles throughout the gastrointestinal (GI) tract facilitate digestion. These contractions are coordinated by rhythmic depolarisations of the smooth muscles, known as slow waves. Impairments to these electrical events are associated with a variety of debilitating motility disorders such as gastroparesis and functional dyspepsia. Gastric pacing is a potential therapy to restore normal slow wave function, similar to how cardiac pacemakers regulate heart rhythms. To investigate slow wave physiology and potential treatments, researchers often rely on acute animal studies under anaesthesia, which provide controlled environments for precise physiological measurements. However, different anaesthetic agents may influence gastric slow wave activity in distinct ways, potentially affecting experimental outcomes. Two common anaesthetic agents are: isoflurane, a widely used volatile anaesthetic, and propofol, a newer intravenous agent gaining broader use. This study compares their effects on gastric slow waves and examines how these differences impact the evaluation of gastric pacing as a therapeutic approach.

Methods: With ethical approval, experiments were conducted on pigs anesthetised with either isoflurane or propofol. To study the slow wave response, surface-contact electrode arrays were placed on the mid-corpus of the stomach, covering both the anterior (front-facing) and posterior (back-facing) surfaces (64 electrodes each side, 5 mm inter-electrode spacing). These surfaces are separated by the greater curvature and the lesser curvature. Following a baseline recording period, pacing was applied using a pulse-width of 400 ms and pulse-amplitudes of 5 mA at 1.1 times the intrinsic slow wave frequency. Slow waves were characterised by their period, amplitude, speed, and spatial propagation patterns, while pacing efficacy was assessed by measuring the rate of spatial entrainment under propofol and isoflurane anaesthesia.

Results: Ordered propagation of slow waves in the antegrade direction was observed across both the anterior and posterior surfaces with 86% symmetry under propofol, while isoflurane resulted in more dynamic propagation patterns and significantly less symmetry (25%, p=0.0187). Baseline slow wave period (18.8±5.1 vs 28.1±14.3 s, p=0.016), amplitude (1.5±0.7 vs 0.7±0.4 mV, p=0.002), and speed (4.4±1.1 vs 3.5±0.7 mm/s, p=0.018) differed significantly between propofol and isoflurane groups, respectively. When evaluating the efficacy of pacing under different anaesthesia, pigs anaesthetised with propofol achieved a spatial entrainment success rate of 83% compared to 57% with isoflurane.

Conclusion: Different anaesthetic agents affect gastric slow waves differently. Propofol preserves more regular activity, while isoflurane is associated with a greater proportion of disordered patterns. This contrast likely stems from their distinct mechanisms of action, with propofol providing targeted sedation and isoflurane exerting broader systemic effects, potentially including on the GI tract. Accounting for these differences is crucial when using acute models to study GI electrophysiology. These differences in anaesthetic effects can be leveraged to study diagnostic and therapeutic interventions where propofol provides a model for normal, ordered slow wave activity, while isoflurane can simulate dysrhythmic states, offering valuable insights into disease mechanisms and potential treatments.

Uterine peristalsis plays an important role in embryo implantation and retention, potentially impacting pregnancy outcomes in assisted reproductive technologies. Previous studies have demonstrated uterine contractions during embryo transfer, but their predictive value for pregnancy outcomes remains unclear. Electrophysiological imaging systems, such as uterine peristalsis imaging (UPI), have allowed non-invasive characterization of uterine peristalsis, though few studies have explored its potential as a predictive biomarker for FET outcomes.

This observational study involved 18 patients undergoing a single blastocyst FET cycle between January to October 2024. Uterine peristalsis was assessed on the day of transfer using a non-invasive electrophysiologic imaging system immediately prior to the FET. Of the 18 patients enrolled, 12 became pregnant and 6 did not following the FET.

Eighteen patients undergoing a single blastocyst FET cycle at an academic fertility clinic were enrolled. Uterine peristalsis was assessed using transabdominal ultrasound for anatomical mapping followed by the application of electrode patches to record electrical signals for up to 30 minutes. UPI software derived peristalsis characteristics (frequency, magnitude, and ratios of C-F and F-C motion). Statistical analysis compared these variables between pregnant and non-pregnant groups using the Mann-Whitney U test.

Both the pregnant (n=12) and non-pregnant (n=6) groups exhibited similar peristalsis frequencies per minute (2.08±0.48 vs 2.07±0.27, p=0.47). No significant differences were observed in the magnitude of uterine peristalsis in either the C-F or F-C directions (p=0.1 and p=0.23, respectively). However, the F-C ratio was significantly higher in the non-pregnant group (0.65±0.13) compared with the pregnant group (0.37±0.12, p<0.02). Conversely, the C-F ratio was significantly higher in the pregnant group (0.63±0.11) compared with the non-pregnant group (0.35±0.14, p<0.02).

Uterine peristalsis characteristics, particularly cervix-to-fundus (C-F) and fundus-to-cervix (F-C) ratios, significantly differ between patients who do and do not achieve pregnancy after FET. The small sample size limits the generalizability of these findings. Additionally, the observational nature of the study does not allow for establishing causality between uterine peristalsis and pregnancy outcomes.

The findings indicate that uterine peristalsis characteristics, particularly the fundus-to-cervix and cervix-to-fundus UP, may serve as a non-invasive biomarker for predicting pregnancy success in FET cycles. Further research is needed to confirm these results and explore clinical applicability.