Accelerated high b-value diffusion-weighted MR imaging via phase-constrained low-rank tensor model

Lianli Liu, Adam Johansson, James M. Balter, Yue Cao, Jeffrey A. Fessler

Abstract: High b-value Diffusion-weighted MRI (DWI) is promising in cancer imaging but suffers from long acquisition time and low signal-to-noise ratio (SNR). We propose a low-rank tensor model that exploits correlation across both diffusion-induced signal decays and neighboring k-space samples, to accelerate the acquisition of DWI using an extended range of b-values (0 s/mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> to 2500 s/mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> ) and limited (orthogonal only) diffusion directions, an imaging scheme that is increasingly used for brain gliomas evaluation. A phase constraint accounts for phase variations between b-values is also applied. Our method integrates parallel imaging and partial Fourier acquisition naturally, and undersamples along phase-encoding direction only. Reconstruction results using both patient and simulated data with an acceleration factor of 8 show improved SNR and reduced aliasing, as compared to parallel imaging only method as well as two other low-rank model-based methods.