A Compressed-Sensing Approach for Super-Resolution Reconstruction of Diffusion MRI

Area of Science:

• Medical Imaging
• Biophysics
• Computational Neuroscience

Background:

• Diffusion MRI (dMRI) is crucial for mapping brain white matter architecture.
• Current dMRI acquisition is time-consuming, limiting spatial resolution.
• Reconstructing high-resolution dMRI is essential for detailed neuroimaging.

Purpose of the Study:

• To develop an innovative framework for reconstructing high-spatial-resolution dMRI from multiple low-resolution (LR) images.
• To reduce dMRI acquisition time while simultaneously increasing spatial resolution.
• To accurately recover very high-resolution diffusion images for in-vivo human brain data.

Main Methods:

• Combines compressed sensing (CS) and classical super-resolution techniques.
• Utilizes subpixel-shifted LR images with down-sampled diffusion directions.
• Employs a sparsifying basis of spherical ridgelets for diffusion signal representation.
• Solves a convex optimization problem using the alternating direction method of multipliers (ADMM).

Main Results:

• Successfully reconstructed high-spatial-resolution dMRI from LR inputs.
• Demonstrated significant reduction in acquisition time.
• Validated the effectiveness on in-vivo human brain data.
• Accurately recovered complex fiber orientations and detailed brain structures.

Conclusions:

• The proposed framework effectively reconstructs high-resolution dMRI.
• This method offers a promising approach to accelerate dMRI acquisition and enhance image quality.
• Enables more detailed analysis of brain white matter architecture.