Compressive Sensing Based Q-Space Resampling for Handling Fast Bulk Motion in Hardi Acquisitions

Area of Science:

• Neuroimaging
• Medical Physics
• Biomedical Engineering

Background:

• Diffusion-weighted MRI (DW-MRI) is crucial for mapping brain connectivity.
• Motion artifacts, particularly fast bulk motion, corrupt DW-MRI data, leading to signal dropout.
• Current methods often discard corrupted data, losing valuable information for analysis.

Purpose of the Study:

• To develop a novel method for resampling corrupted slices in DW-MRI data caused by fast bulk motion.
• To improve the reconstruction of diffusion orientation distribution functions (ODFs) from under-sampled and corrupted measurements.
• To mitigate the loss of data associated with traditional gradient exclusion methods.

Main Methods:

• Utilized compressive sensing-based reconstruction techniques for ODFs.
• Employed Simple Harmonic Oscillator based Reconstruction and Estimation (SHORE) basis functions for analytical ODF modeling.
• Applied the proposed resampling strategy to simulated intra-gradient motion data and real DW-MRI datasets.

Main Results:

• The proposed q-space resampling method effectively recovers corrupted slices affected by fast bulk motion.
• Demonstrated superior performance compared to existing state-of-the-art resampling techniques and gradient exclusion.
• Successfully preserved essential gradient information for subsequent diffusion MRI reconstruction.

Conclusions:

• The compressive sensing-based resampling strategy offers a robust solution for handling motion artifacts in DW-MRI.
• This approach enhances the integrity of diffusion data, particularly in challenging patient populations.
• Enables more accurate and complete brain connectivity analysis from DW-MRI.