Progress Towards Model-Based Estimation of the Cardiac Electromechanical Activity from ECG Signals and 4D Images

We present recent advances on a 3D numerical modeling of the myocardium which couples electrical and biomechanical models. The long-term objective is to simulate a realistic contraction of the heart from both electrical measurements (typically the ECG) and geometrical measurements (typically provided by medical imaging). This realistic contraction should provide useful quantitative parameters for the diagnosis and also for guiding some new forms of therapy. Our modeling is based on a multi- scale analysis ranging from microscopic to macroscopic scales, and integrates a priori information on the overall geometry and on the fiber directions extracted from specific medical imaging techniques (e.g. dtMRI). The FitzHugh-Nagumo equations are solved along with a constitutive law based on the Hill-Maxwell Rheological model, on which a data assimilation analysis is done. In medical image analysis, we believe that this new generation of physics-based deformable models will be useful to provide a more robust quantitative interpretation of temporal series of cardiac images. Resume. Nous presentons ici nos avancees sur la modelisation du myocarde couplant des modeles ´ et biomecaniques. L'objectiflong terme est de simuler la contraction du cœurpartir de mesures ECG et d'images medicales, afin de fournir des outils d'aide au diagnostic. Notre modele se base sur unemultiechelle des phenomenes et des donnees anatomiques extraites d'images medicales. La partierepose sur le modele de FitzHugh-Nagumo et la partie mecanique reprend le formalisme de Hill-Maxwell avec une nouvelle loi de comportement, sur laquelle de l'assimilation de donnees est realisee. En imagerie medicale, nous pensons que ce nouveau type de modeles devrait permettre une interpretation plus robuste des sequences temporelles.