Robust Volumetric Diffusion-Weighted MRSI via Time-Resolved Phase Reconstruction and Correction

Motivation: To address the long-standing phase correction challenge and enhance the robustness for diffusion-weighted MRSI. Goal(s): To correct the significant phase variations due to macroscopic and microscopic motions in in vivo diffusion-weighted MRSI acquisition. Approach: We developed a novel fast diffusion-weighted MRSI sequence integrating time-resolved, sparsely sampled, volumetric phase navigators, a subspace-based phase image reconstruction, and a sensitivity-encoded phase-corrected reconstruction. The corrected diffusion-weighted MRSI data were processed by state-of-the-art subspace-based spatiospectral processing methods. Results: Improved data quality, diffusion-weighted spatiospectral reconstruction and metabolite-specific diffusion parameter estimation achieved by the proposed method are demonstrated using in vivo data. Impact: A novel integrative acquisition and reconstruction solution for robust, phase-corrected 3D in vivo diffusion-weighted MRSI was presented, an important step towards developing diffusion-weighted MRSI for its translation to quantitative, molecule-specific microstructural imaging.