Restriction-Weighted Q-Space Trajectory Imaging (ResQ): Toward Mapping Diffusion-Time Effects With Tensor-Valued Diffusion Encoding in Human Prostate Cancer Xenografts.

Tensor-valued diffusion encoding employs gradient waveforms that enable unique sensitivity to microstructural features of tissue, but the interpretation of signal and parameters may be confounded by diffusion-time dependence. We introduce a framework for restriction-weighted q-space trajectory imaging (ResQ) that incorporates diffusion-time effects via the restriction-weighting tensor, and we evaluate it in a longitudinal study of prostate cancer xenografts treated by external radiotherapy. We proposed a novel gradient waveform design for tensor-valued encoding with controlled restriction weighting and applied a set of four waveforms at a 9.4 T preclinical MRI system. Mice were inoculated with human prostate cancer cells (LNCaP) and assigned to groups that were untreated controls or treated by external beam irradiation. ResQ produced parameters that describe the diffusion process in terms of the mean diffusivity (D), isotropic diffusional variance (VDi), and microscopic diffusion anisotropy (VDa) as well as their diffusion-time dependence (ΔD, ΔVDi, ΔVDa). Analyses were performed to characterize parameters longitudinally and across groups. To highlight the consequences of ignoring restriction effects, we compared ResQ to analogous parameters estimated by q-space trajectory imaging (QTI). ResQ revealed clear diffusion-time dependence across all tumors, with significant longitudinal differences between treated and untreated groups, most prominent in D, ΔD, and VDi. The ResQ signal representation captured the signal dynamics, whereas QTI did not. Neglecting diffusion-time dependence in QTI led to substantial parameter bias, most notably a pronounced overestimation of microscopic diffusion anisotropy. Diffusion-time effects are non-negligible in prostate cancer and must be considered when using tensor-valued diffusion encoding. The ResQ framework enables controlled restriction weighting and improved interpretability of diffusion MRI parameters compared with approaches that ignore the effects of restriction. This provides a more principled approach for tensor-valued diffusion encoding and may enable novel imaging biomarkers that disentangle diffusivity, isotropic diffusional variance, microscopic anisotropy, and their diffusion-time dependence.