Diffusion-weighted GRASE sequence with 3D navigator for high-resolution time-dependent diffusion MRI in the human cortical gray matter

Abstract

Purpose: Oscillating gradient (OG) and pulsed gradient spin-echo diffusion MRI (dMRI) allows the acquisition of diffusion-weighted signals at varying diffusion time (t_d) and thus enables characterization of microstructure in complex tissues. This study proposed an advanced three-dimensional (3D) navigator-based diffusion-weighted 3D gradient spin-echo (DW-GRASE) sequence, specifically designed for high-resolution whole-brain oscillating gradient spin-echo and pulsed gradient spin-echo acquisition.

Methods: We developed a 3D navigator for phase correction between multiple shots of the 3D DW-GRASE sequence. Oscillating and pulsed gradient dMRI were acquired at 1.5-mm isotropic resolution with oscillating frequencies at 50 and 25 Hz and t_d of 20, 30, and 40 ms. The performance of two-dimensional and 3D navigator strategies was compared based on the ghost-to-signal ratio. Then, we evaluated t_d-dependency in three cortical regions, as quantified by diffusion dispersion exponent (θ) based on the power-law relationship between t_d and apparent diffusion coefficient.

Results: dMRI obtained using the 3D navigator method showed significantly lower ghost-to-signal ratio than the two-dimensional navigator, indicating high performance in correcting phase errors between shots. The θ values obtained from the 3D navigator-based 3D DW-GRASE sequence were higher in the sensory and motor regions than in the higher-order region, indicating more complex microstructures in the higher-order cortex.

Conclusion: The 3D navigator-based oscillating gradient DW-GRASE sequence effectively corrected phase errors and demonstrated the potential to achieve high-resolution t_d-dependent dMRI for revealing microstructural properties in the human cortex.

Keywords: 3D GRASE; 3D navigator; diffusion MRI; diffusion dispersion exponent; phase error; time dependence.