The effects of leaflet material properties on the simulated function of regurgitant mitral valves

TL;DR

This study tackles the challenge of noninvasively obtaining patient-specific leaflet material properties by evaluating how variations in tissue extensibility influence FE predictions of mitral valve function and mechanics. Using 3D image-derived MV geometries, an isotropic Lee-Sacks constitutive law, and a novel automated ROA quantification, the authors generate normal and three regurgitant valve morphologies with five material variants, coupled with traditional and statistical uncertainty analyses. They find that the relative ordering of functional and mechanical metrics is preserved when tissue extensibility varies within approximately $15\%$, suggesting geometry is the dominant determinant of immediate and long-term valve behavior under material-property uncertainty. These results support the use of image-derived FE modeling to qualitatively compare valve repairs, particularly in children, while highlighting the need for future validation and more precise tissue-property data.

Abstract

Advances in three-dimensional imaging provide the ability to construct and analyze finite element (FE) models to evaluate the biomechanical behavior and function of atrioventricular valves. However, while obtaining patient-specific valve geometry is now possible, non-invasive measurement of patient-specific leaflet material properties remains nearly impossible. Both valve geometry and tissue properties play a significant role in governing valve dynamics, leading to the central question of whether clinically relevant insights can be attained from FE analysis of atrioventricular valves without precise knowledge of tissue properties. As such we investigated 1) the influence of tissue extensibility and 2) the effects of constitutive model parameters and leaflet thickness on simulated valve function and mechanics. We compared metrics of valve function (e.g., leaflet coaptation and regurgitant orifice area) and mechanics (e.g., stress and strain) across one normal and three regurgitant mitral valve (MV) models with common mechanisms of regurgitation (annular dilation, leaflet prolapse, leaflet tethering) of both moderate and severe degree. We developed a novel fully-automated approach to accurately quantify regurgitant orifice areas of complex valve geometries. We found that the relative ordering of the mechanical and functional metrics was maintained across a group of valves using material properties up to 15% softer than the representative adult mitral constitutive model. Our findings suggest that FE simulations can be used to qualitatively compare how differences and alterations in valve structure affect relative atrioventricular valve function even in populations where material properties are not precisely known.

Figures (17)

• Figure 1: Conceptual future application workflow. Potential future application workflow using image-derived simulation to explore and optimize the optimal repair for an individual patient. The workflow goes as follows 1) create a patient-specific 3D valve model from 3D image, 2) perform variations of virtual intervention to the valve model, and 3) identify the optimal repair by comparing the biomechanical and functional metrics for each option.
• Figure 2: Mitral valve geometry and anatomy. (A) Open mitral valve with annulus, free edge, leaflets, and chordae tendineae shown; (B) Closed mitral valve with regions of the anterior leaflet (A1, A2, A3) and posterior leaflet(P1, P2, P3) labeled.
• Figure 3: Overview. We created models of one normal (non-regurgitant) and three regurgitant MV with different mechanisms of regurgitation ( posterior leaflet tethering, posterior leaflet prolapse in the P2 region, and symmetric annular dilation). For the three dysfunctional morphologies (Tethered, P2 prolapse, Annular dilation) we created morphologies with two degrees of regurgitation (moderate and severe). We used these MV morphologies to examine the effects of tissue extensibility and the individual material coefficient on the mechanical and functional metrics. Visualization of chordae omitted for clarity.
• Figure 4: ROA computation procedure. (A) Create a closed contour near the valve opening. (B) SlicerHeart automatically generates a continuous surface. The red region indicates the area where the continuous surface is in contact with the leaflet surface; blue indicates the potential orifice surface. (C to D) the potential orifice surface gradually descends toward the orifice opening in the valve model by a user-defined number of shrink-wrapping iterations. (E) Generate 400 rays (red lines) over the potential orifice surface to identify the streamlines that pass through the valve without intersecting the leaflets. (F) Compute the effective ROA by summing the areas around the rays that do not intersect the leaflet surface.
• Figure 5: Material variants. We derived five material variants, each representing a gradual increase of tissue extensibility relative to the baseline mitral model, by performing a uniaxial test using FE analysis. The material constants are shown in (A), their corresponding stress-strain curves are shown in (B), and the percentage difference of the stretch ratio, $\lambda$, between each material variant and the mitral model (C).