High-fidelity, accelerated whole-brain submillimeter in vivo diffusion MRI using gSlider-spherical ridgelets...

Area of Science:

• Neuroimaging
• Medical Physics
• Biomedical Engineering

Background:

• Diffusion MRI (dMRI) is crucial for in vivo human brain imaging.
• Achieving submillimeter resolution requires long acquisition times, limiting clinical utility.
• Existing techniques face challenges in balancing speed, resolution, and data quality.

Purpose of the Study:

• To develop an accelerated, robust, and accurate diffusion MRI acquisition and reconstruction technique.
• Enable submillimeter whole-human-brain in vivo scanning on clinical MRI scanners.
• Reduce the total acquisition time of conventional gSlider technique.

Main Methods:

• Extended the gSlider technique by undersampling in q-space and RF-encoding space, termed gSlider-SR.
• Compensated for undersampling by exploiting data redundancy using spherical ridgelets (SR) and enhancing signal-to-noise ratio.
• Validated using Monte Carlo simulations and in vivo human brain dMRI data on a Siemens Prisma 3T scanner.

Main Results:

• gSlider-SR successfully reconstructed high-quality dMRI data at various acceleration factors, preserving signal and angular information.
• Achieved 860 μm resolution whole-brain dMRI with 64 diffusion directions in 10 minutes, comparable to conventional gSlider with four averages (1 hour 20 minutes).
• Demonstrated an eight-fold reduction in scan time for in vivo human brain data.

Conclusions:

• gSlider-SR enables whole-brain high angular resolution dMRI at submillimeter resolution with drastically reduced acquisition time.
• The proposed scheme is feasible for use on existing clinical MRI scanners.
• Accelerated dMRI acquisition and reconstruction techniques are vital for advancing neuroimaging research and clinical applications.