Oscillating Gradient Spin Echo (OGSE): A Study of Short Diffusion Time Effects in Human Brain at 3T

Abstract

Diffusion-weighted magnetic resonance imaging (DWI) is a non-invasive MRI technique that is sensitive to the diffusion of water molecules within the body. Its ability to encode water displacements enables it to detect changes in neural microstructure, such as those due to normal healthy aging or even those caused by pathological conditions. Diffusion MRI sequences will be less or more sensitive to diffusion depending upon the time that they encode or measure these displacements, known as the diffusion time. For instance, the oscillating gradient spin echo (OGSE) diffusion sequence encodes water diffusion within the frequency spectrum of its gradient oscillations. As the frequency is increased, the diffusion time decreases and water molecules will reach less cellular boundaries over larger spatial scales. With less restrictions to diffusion, measurable changes in diffusion parameters can be detected, such as an increase in mean and radial diffusivities (MD and RD, respectively). Hence, OGSE can unmask spatial-dimension tissue differences (e.g. such as axon diameters) and achieve unique imaging contrast mechanisms related to cell size, as found previously in animal models. However, translating OGSE to human applications is highly challenging due to hardware constraints (such as limited gradient strength) in clinical scanners. Consequently, there are only a few OGSE studies in healthy human brain, using frequencies in the range of f = 18-63 Hz. In this work, OGSE was evaluated in human brain on a 3T clinical MRI system with oscillation frequencies of f = 40 – 50 Hz. The acquired images provided adequate data quality and greater OGSE MD values relative to the MD from the conventional diffusion sequence pulsed gradient spin echo (PGSE). Nonetheless, raw OGSE and PGSE scans demonstrated image artifacts such as Gibbs ringing (GR) that affected the accurate estimation of diffusion metrics, necessitating image acquisition optimization. Fluid-attenuated inversion-recovery (FLAIR) was implemented for the first time with OGSE in acquisition to suppress the sharp signal intensity transitions at cerebrospinal fluid (CSF)/tissue boundaries causing the ringing. FLAIR was found effective in preventing GR as compared to remedial post-acquisition correction methods, as it substantially increased image quality in OGSE/PGSE DWIs and provided homogeneous MD maps. Region-of-interest analysis was then performed on several white matter tracts and two deep gray matter structures in eight subjects on OGSE FLAIR - PGSE FLAIR diffusion maps. Results from the OGSE-PGSE difference maps showed significantly elevated MD and RD with shorter diffusion time in the corticospinal tract, superior longitudinal fasciculus, and posterior limb of the internal capsule, that was greater than in the other white matter tracts and both gray matter regions. OGSE FLAIR, although characterized by lengthy scan times, proved a reliable and effective method to investigate potential axon-scale differences in healthy white matter, with potential to explore changes in tissue microstructure in the mechanisms of disease.