Reducing acquisition time for microstructure imaging with spatially-regularized global optimization

Microstructure imaging with diffusion MRI (dMRI) is usually based on fitting models to the data using voxelwise nonlinear procedures that (i) are time consuming and (ii) need long acquisitions. Recently, we introduced the AMICO framework to accelerate the fit; here, we propose an extension to enable faster acquisitions, too. Our formulation allows exploiting information from neighboring voxels that is not accessible with existing approaches.