Restriction-weighting in tensor-valued diffusion encoding: A novel experimental framework demonstrated in human prostate cancer xenograft

Motivation: Restricted diffusion effects in tensor-valued encoding can lead to parameter error due to diffusion time-dependence. Goal(s): Accurate interpretation of diffusion process by joint quantification of restricted diffusion, microscopic diffusional variance/anisotropy, and their covariances. Approach: We propose a novel diffusion encoding scheme and an extended signal representation to capture both Gaussian and restriction effects. We demonstrate the method in a human prostate cancer xenograft in vivo (mouse model). Results: The analysis reveals prominent effects of restricted diffusion and time-dependence in all parameter maps, and the covariances terms were found to be non-negligible. Impact: We propose an experimental design and analysis of tensor-valued diffusion encoding that incorporates effects of restricted diffusion. The results shed light on confounding effects of restricted diffusion in tensor-valued encoding experiments with mixed encoding frequencies.