Accuracy and precision in super-resolution MRI: Enabling spherical tensor diffusion encoding at ultra-high b-values...

Area of Science:

• Medical Imaging
• Biophysics
• Neuroscience

Background:

• Diffusion MRI (dMRI) provides insights into tissue microstructure.
• Acquisition challenges include low signal-to-noise ratio (SNR) at high resolution and diffusion encoding strengths, leading to signal inaccuracies like positive bias.
• Super-resolution (SR) techniques offer a potential solution by reconstructing high-resolution images from low-resolution, high-SNR data.

Purpose of the Study:

• To introduce and evaluate a super-resolution reconstruction framework for dMRI.
• To investigate the impact of SR on signal accuracy and precision in dMRI.
• To demonstrate the utility of SR for advanced dMRI contrasts in the human brain.

Main Methods:

• Development of a super-resolution reconstruction framework tailored for dMRI data.
• Validation using phantom experiments and numerical simulations to assess accuracy and precision.
• Application of the SR framework to in vivo human brain data with advanced diffusion encoding (spherical b-tensor, ultra-high b-values).

Main Results:

• Super-resolution reconstruction significantly improves dMRI signal accuracy by optimizing the resolution-SNR trade-off before noise floor effects.
• Precision gains from SR are substantial but depend on sacrificing some spatial resolution.
• SR enables high-contrast imaging for diffusion restriction in all directions, previously limited by low SNR or resolution.

Conclusions:

• Super-resolution reconstruction is a valuable tool for improving dMRI accuracy and enabling advanced imaging contrasts.
• The technique allows for more beneficial trade-offs between spatial resolution and diffusion encoding strength.
• SR facilitates investigations of complex tissue microstructures with improved SNR and resolution compared to conventional methods.