Compressive MRI quantification using convex spatiotemporal priors and deep encoder-decoder networks

•Fitting to the Bloch manifold via dictionary matching is computationally intensive.•Deep networks can compactly do this via hierarchical partitioning and multiscale approximations.•We propose a 2-stage dictionary-matching-free pipeline for multi-parametric quantitative MRI.•Convex spatiotemporal re...

Full description