Devon M. Middleton, Abstract Co-Author: Nothing to Disclose

Scott Hunter Faro MD, Presenter: Nothing to Disclose

Mary Jane Mulcahey, Abstract Co-Author: Nothing to Disclose

Feroze B. Mohamed PhD, Abstract Co-Author: Nothing to Disclose

Spinal cord diffusion tensor imaging (DTI) is a relatively new field of research that may lead to better understanding of physiologic changes in many spinal cord pathologies. Physiological motion can create problems in DTI of the spinal cord. Spinal cord DTI is complicated by oscillation and pulsation of the cord, and by noise introduced by cardiac and respiratory motion. If motion is not corrected for, results of DTI analysis are impacted, including, false decrease in white matter fractional anisotropy (FA), poor delineation of the cord/CSF interface, and poor gray/white matter differentiation. The purpose of this work was to show efficacy of rigid body motion correction techniques in improving cord/CSF conspicuity and gray/white matter definition.

DTI data was collected for five healthy pediatric subjects on a Siemens Verio 3T MRI scanner using an inner field of view EPI sequence with 2DRF excitations. Imaging parameters were: TE = 110 ms, TR = 7900 ms, Voxel 0.8 x 0.8 x 6 mm3, 20 diffusion directions, 3 av, 6 b0, b = 800 s/mm2. Prior to correction, a mask was applied to the center of the image to eliminate data beyond the spinal canal. First, b0 acquisitions were co-registered by rigid body transformation and averaged to create a mean b0 using SPM8. All DW images were registered to the mean b0 image using a rigid body registration method implemented in the ACID toolbox. For each subject, ROIs were drawn on FA maps for 3 adjacent axial slices to include lateral and posterior white matter and exclude gray matter. FA values were examined for both corrected and uncorrected images.

Clear improvements were visible in cord/CSF delineation and in gray/white matter definition in FA maps after motion correction. In some cases, improvement was dramatic, making initially unusable data clear. In all cases, corrected images showed higher FA values for white matter ROIs compared with uncorrected, ranging from 3-38% increases.

Rigid body motion correction led to an increase in FA values for white matter ROIs in the corrected data of the pediatric spinal cord. Additionally, there was improved definition of cord/CSF interface and gray/white matter differentiation.

Rigid body motion correction showed an increase in FA values for white matter ROIs and improved cord/CSF interface and gray/white matter differentiation in DTI of the healthy pediatric spinal cord.

Middleton, D, Faro, S, Mulcahey, M, Mohamed, F, Improving Fractional Anisotropy Measurements and Gray Matter/White Matter Differentiation in DTI of the Pediatric Spinal Cord Using Rigid Body Motion Correction.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14019442.html