Optimizing Acquisition and Post-Processing Pipeline for Motion-Robust Diffusion Weighted Imaging of the Liver

Motivation: M1-optimized diffusion imaging (MODI) improves motion robustness but increases echo time and has occasional residual motion-induced artifacts. Goal(s): To optimize the acquisition and post-processing pipeline for motion-robust diffusion MRI. Approach: Monopolar and varying M1-value MODI waveforms were acquired for 7 volunteers and reconstructed using unweighted and weighted averaging across repetitions and diffusion directions. ADC maps were created for each combination, and measured in right and left liver lobes to assess interlobar agreement. Results: Interlobar bias was reduced from 0.66x10-3mm2/s (monopolar acquisition with unweighted averaging) to 0.02x10-3mm2/s (MODI with M1=0.63s/mm acquisition with weighted averaging). Impact: Advanced acquisition and post-processing pipelines, including jointly optimized motion-robust diffusion imaging (MODI) and weighted averaging, may improve the quantitative performance of ADC in the liver. This improved quantification may enhance the detection, staging, and treatment monitoring of liver disease.