Predictive approaches

Prediction of alterations evolution. Statistical models will be developed to capture the evolution of vessel alterations describing the aging process. Their goal is to predict the degradation evolution of the vascular system from extracted features for a given individual at a given time. To reach this objective, both parametric models and machine learning methods will be used and merged to provide a global model.

Prediction of vascular function. Capitalizing of previously developed tools, 3D mesh networks fitting the aortic lumen and wall will be produced. These envelopes will be further used as deformable conduits for computational fluid dynamics simulations. These calculations will benefit from the data gathered on vascular stiffness and flow velocities recorded for each individual. These computations will rely on parallel computing (HPC Romeo center) as a full 3D finite element fluid-structure model is expensive to solve. These computations for vessel wall of different ages will allow to compute the impact of aging on the blood flow.