oa Integration of Microstructural Architecture of the Mitral Valve into an Anatomically Accurate Finite Element Mesh

Although mitral valve (MV) repair initially restores normal leaflets coaptation and stops MV regurgitation, in long term it can also dramatically change the leaflet geometry and stress distribution that may be in-part responsible for limited repair durability. As shown for other collagenous tissues, such changes in geometry and loading reorganize the fiber architecture. In addition, MV interstitial cells may also respond to the altered stress by reducing biosynthetic function, which would affect the load-bearing capabilities of MV and its long-term durability. Thus, investigating the repair-induced MV stress and the concomitant microstructural alterations is a key step in assessing the repaired valve durability. Finite element models have been widely used for stress analysis of the mitral valve. Most of these models, however, have employed only basic constitutive models and above all ignore the complex microstructure of the MV. In addition, the geometry of the valve is usually simplified. Thus, in this work we developed a method to obtain accurate geometrical model of the ovine MV and quantify its fiber structure for the purposes of developing high fidelity computational meshes of the MV. To obtain an accurate geometry of the MV, microcomputed tomography (micro-CT) was used. The entire heart was scanned via a SIEMENS Inveon CT scanner. Three-dimensional scans were segmented semi-automatically using ScanIP segmentation software. The 3D positional data of the fiducial markers were also obtained via ScanIP masks generated by using gray-scale threshold of the CT scans. The segmented geometry was then converted to finite element meshes using ScanIP mesh free mesh generator scheme. Next, the anterior leaflet was then dissected and prepared for measurements of its fiber alignment. The positional data of each point on the accurate mesh was then projected onto the 3D marker mesh. By using a computational domain, the projected point was mapped back to the 2D flattened surface. In addition to mapping, the current method can be used to estimate the changes in connective tissue structure with deformation. This is done by for each point on the valve surface using the local right Cauchy strain tensor C using an in-plane convective curvilinear coordinate system to convect the local fiber orientation to predict the current fiber alignment. To conclude, a robust technique to quantify and map the fibrous microstructure of the MV anterior leaflet to anatomically accurate 3D MV shape derived from micro-CT imaging was developed. The method provides a framework for development of anatomically and micro-structurally accurate finite element models of MV using our tissue structure-based models. It can also be used as a means to validate predicted changes in fibrous structure due to altered stress following surgical interventions.