Project Summary
We propose to optimize and validate two novel diffusion MRI models/methods that have direct clinical
relevance for cervical spinal cord evaluation in health and multiple sclerosis (MS): Neurite Orientation
Dispersion and Density Imaging (NODDI) and Spherical Means Technique (SMT). While diffusion tensor
imaging (DTI) has existed for 20 years, advanced biophysical models for evaluating neurological disease in the
CSC are lacking. Advanced diffusion MRI can extract indices related to neural architecture and axonal loss, yet
evaluating the pathological substrates of MS (specifically axonal loss) in the CSC is undertested and questions
remain if the models as-developed are relevant for pathology. Lastly, it is not clear if advanced diffusion MRI
offers greater clinical value over DTI. We address the current knowledge gap in human CSC diffusion MRI by
optimizing and evaluating two clinically-approachable diffusion techniques: NODDI and SMT in healthy
volunteers and patients with relapsing-remitting MS (RRMS) to 1) study lesion and normal appearing white
matter (NAWM) in comparison with conventional DTI (assessing value), lesion burden, and atrophy (reflecting
axonal loss) and 2) to assess the sensitivity of diffusion MRI to tissue change over time. In MS, spinal cord
health is integral to neurological function, yet current studies rely on identifying lesions and/or tissue atrophy;
the biological substrates of CSC tissue damage are poorly characterized, and their relationship to neurological
function is weak. Advanced diffusion MRI provides estimates of axonal volume, cellular inflammation, and
neurite dispersion and may provide greater specificity than DTI for microstructural changes in the CSC
throughout MS evolution. However, advanced diffusion MRI has only recently been explored due to lack of
CSC-optimized acquisitions and models that account for pathology, which we show are surmountable. We will
test the hypotheses that NODDI and SMT diffusion MRI, can 1) detect sub-radiological axonal pathology in MS
(CSC areas devoid of lesions), 2) offer improved value and specificity over conventional DTI, and 3)
characterize axonal-sensitive indices longitudinally concomitant with neurological deterioration. We optimize,
and acquire NODDI and SMT data in addition to DTI, T2-, T2*-, T1 MRI in the CSC of patients with RRMS. We
published NODDI and SMT in the CSC in a small cohort of RRMS patients, but now evaluate the value that
advanced diffusion modeling in the CSC can add to the clinical assessment of MS patients. As in the brain,
NODDI and SMT can be acquired in a reasonable exam time, but are untested for spinal cord pathology in MS.
If successful, we will offer clinically-relevant, optimized acquisition and analysis tools for the application of
advanced diffusion MRI in CSC pathology in comparison with clinical radiological standards. Alternatively, we
will solidify the importance (and provide optimized CSC sequences) for rapid, conventional CSC DTI for clinical
deployment. A byproduct is the evaluation of alternative models in pathology, which has not been tested before
and have direct benefit to understanding other spinal cord diseases.