Computational Cardio-Vascular
Cardiovascular diseases and their acute and chronic consequences on heart and brain are responsible for more than 40% of deaths in the developed countries. In this context, we develop multi-modal approaches based on innovative acquisition techniques combining spectral photon counting CT scanner and PET-MR along with machine learning techniques for multiple data integration and population analysis
  • Spectral photon counting CT scanner - SPCCT: Our goal is to detect, quantify and monitor in vivo the evolution of atheroma, myocardial infarction, and wall remodeling using SPCCT, a new type of imaging that improves the sampling of X-ray spectral information.
  • PET-MR and MRI: We setup a comprehensive study of the multi-factorial progression of vascular disease. We thus improve data acquisition and reconstruction strategies in order to obtain in vivo measurements of blood velocity, wall motion and inflammation.
  • DTI and SPCCT: Getting simultaneously insights into fiber structure and material distribution has never been studied before and presents the potential to create new biomarkers for the human heart. We thus investigate advanced deep learning-based methods and their applicability to in vitro and in vivo data.
  • Diagnosis and prognosis from routine imaging modalities: The richness of cardiac data from existing protocols in clinical routine is still under-exploited. We thus investigate: i) the increase in robustness and reliability the current measurements extracted from conventional imaging systemsii) the integration of complex descriptors thanks to representative learning solutions; iii) the exploitation of advanced statistical models to improve diagnosis and prognosis.