Conference Agenda

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.