GENERATION OF A SYNTHETIC DATABASE OF HEART MODELS TO EVALUATE STRAIN-BASED INFARCT DETECTION ALGORITHMS

International audienceAccurate quantification of myocardial viabilityis essential to determine the best clinical treatment for heart disease patients [1]. A decline in myocardial viabilityinfluences the structural and functional properties of the heart.Regional myocardial function can be assessed via strain analysisextracted fromtagged or tissue tracking MR imaging. However, further processing is required to localize infarct based on these strains.Our previous study showed promising results in using strain-based parameters to detect myocardial infarcts in the diastolic phase [2].In this study, we introduce amethodology to design a large synthetic database of infarcted hearts based on existing healthy cases with distinct geometries, enabling ground truth data. This database will be used to evaluate the performance of the algorithms developed to locate infarcts in the left ventricle (LV).1. Bhat et al. Cardiol Res Pract: 1020818. 2016.2. Rumindo et al. CMBBE 2016