Fifty-sixth annual meeting of the American association of physicists in medicine
SU-E-I-50: Ultrahigh Resolution Reduced Field-Of-View Diffusion Tensor Imaging in the Spinal Cord
Diffusion Tensor Imaging (DTI) is a well-established MRI imaging modality for applications in the brain. The typically attainable imaging resolution for clinical DTI sequences is insufficient to image small lesions and assess injury with microstructural changes in the spinal cord (SC). Attempts to reduce the field-of-view (FOV) for increasing the imaging resolution have been made using rectangular FOV geometries that work well to efficiently image the sagittal plane. The highest structural details in the SC are however in the axial plane, most efficiently covered by a circular FOV. The aim of this study was to increase the axial DTI resolution in the SC to distinguish the dorsal spinal white-matter bundles within clinically acceptable acquisition times.
A multishot reduced FOV DTI technique named eZOOM that utilizes spatially selective radio-frequency (RF) pulses and spiral readout has recently been developed by the author (Gaggl, MRM 2014) for application in the brain. In this study the method was adapted to the unique geometry of multichannel spinal RF coil arrays, with improved correction of blurring and misregistration and including peripheral gating to minimize cerebrospinal fluid flow-artifacts. Several diffusion metrics including the novel normalized axial diffusivity DAn=λ1/∑λ (Gaggl, ASFNR 2013) were calculated and compared to current methods.
Sub-millimeter square-voxel imaging resolutions with image voxel volumes of approximately 1mm3 for a circular axial 8-cm FOV and 15 diffusion weighting gradient directions were achieved in human for multislice image stacks in the SC with total imaging times of less than 30 minutes at 3T, enabling the separation of important spinal fiber tracts.
The presented technique improves imaging resolution for DTI in the spinal cord compared to current clinical sequences while allowing to prescribe imaging geometries that excite only a small axial circular FOV in and around the spinal cord and reducing blurring and artifacts.