Diffusion restriction in the human spinal cord characterized in vivo with high b-value STEAM diffusion imaging

Restricted or hindered motion of water across axonal membranes as characterized with diffusion-weighted (DW) imaging may be a potential marker of axonal damage in white matter (WM) injury due to trauma, neurodegeneration, or other causes. This study sought to determine whether high b-value DW imaging with a stimulated echo (STEAM) sequence could improve the spatially resolved assessment of tissue architecture in the human spinal cord in vivo. Diffusion times from 76 ms to 1000 ms and b-values of up to 14,750 s/mm2 were used to acquire axial DW images in six healthy volunteers, and four additional healthy volunteers were studied with a protocol focused on high b-value, higher-resolution imaging. Mono-exponential, diffusional kurtosis, and mono-exponential with an additive constant (MEC) models were fit individually to diffusion decay curves obtained at different diffusion times. Diffusion restriction, characterized with the diffusional kurtosis and MEC models, was measured more precisely using higher b-value ranges. DW images at high b-value and fitting parameters using the large range of b-values available at the diffusion time of 1000 ms demonstrated signal and restriction differences between gray and white matter and even across white matter regions. These white matter differences may reflect variations in axonal density, diameter, or alignment. We conclude that high b-value DW imaging with a STEAM sequence on a conventional clinical scanner can provide accurate measures of diffusion hindrance and restriction in human spinal cord in vivo.