Conference Agenda

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)