Reconstruction of blood flow velocity with 4D MRI and physics enforced deep learning

Our goal for this project is to improve blood velocity field reconstruction from undersampled k space 4D MRI data, both in terms of computation time, image quality, and velocity quantification. We intend to use new deep learning approaches with physically based regularization to increase the spatio-temporal resolution and decrease the calculation time. Our aim is to combine the knowledge of physical laws with deep learning techniques order to capture the fluid patterns.  The efficiency of the frameworks will be demonstrated with numerical examples with non-stationary velocity fields calculated with realistic simulations and test the methods on sparse and noisy k-space signals obtained with real data from others.

Keywords Blood flow reconstruction, 4D MRI, deep learning, physics informed machine learning