My research focuses on the characterization of diseases from medical imaging populations, in particular for cardiac imaging. I also lead several initiatives to promote engineers-cardiologists interactions and boost the value of such tools for clinicians.

On the methodological side, I develop new statistical and computational approaches to represent medical imaging populations, based on statistical atlases and machine learning. This also covers pre-processing (segmentation, tracking, and mesh/image registration) and validation (generation of synthetic databases).

On the applicative side, my work aims at better understanding disease apparition and evolution, for computer-aided diagnosis purposes. I focus on cardiac function and in particular 2D/3D myocardial shape and deformation, with widespread imaging modalities such as 2D/3D echocardiography and magnetic resonance.