Accurate super-resolution of diffusion MRI data at ultra-strong gradients and varying diffusion time using Image Quality Transfer

Motivation: Diffusion-weighted imaging (DWI) is a powerful technique for investigating brain microstructure and obtaining quantitative measures of microstructural features. However, current resolution limits biological accuracy. Goal(s): Image Quality Transfer (IQT) was adapted to use spherical harmonics coefficients to enhance DWI resolution acquired with ultra-high b-values and multiple diffusion times. Approach: IQT performance was assessed in two DWI acquisitions, then enhanced-resolution DWI-data was fitted with advanced biophysical models to demonstrate the impact of resolution enhancement, alongside quantitative and histological accuracy. Results: IQT reconstructed with high accuracy enhanced-resolution DWI-data, whose fitting provided quantitative and histological accurate maps, highlighting IQT benefits for research and clinical applications. Impact: We demonstrated IQT effectiveness in advanced DWI acquisitions with ultra-high b-values and multiple diffusion times when adapted with suitable signal representation. This approach advances the precise study of brain microstructure, overcoming current DWI limitations in both clinical and research settings.