Image Reconstruction from Accelerated Slice-Interleaved Diffusion Encoding Data

We present a reconstruction scheme for diffusion MRI data acquired using slice-interleaved diffusion encoding (SIDE). We show that, when combined with multi-band imaging, the method is capable of reducing the amount of data that needs to be acquired by as much as 25 times, therefore remarkably speeding up acquisition and making high angular resolution diffusion imaging much more feasible, particularly for pediatric, elderly, and claustrophobic patients. In contrast to the conventional approach of acquiring a full diffusion-weighted (DW) volume for each diffusion wavevector, SIDE acquires for each repetition time (TR) a volume consisting of interleaved slice groups, each corresponding to a different diffusion wavevector. This allows SIDE to rapidly acquire information associated with a larger number of wavevectors within a short period of time. We formulate the inverse problem involved in recovering the full DW images as a constrained variational problem regularized by multidimensional total variation. The problem can be solved efficiently using the alternating direction method of multipliers (ADMM). Experiment results based on SIDE data of adults indicate that DW images can be recovered with high fidelity despite high undersampling for multifold acceleration.