Reporting Quantitative Microstructural MRI Metrics to Stage and Monitor Multiple Sclerosis

PROJECT SUMMARY At least 400,000 people in the United States are affected by Multiple sclerosis (MS), which is the most prevalent chronic inflammatory disease of the central nervous system (CNS). The current non-invasive MRI gold-standard of MS characterization is through volumetric biomarkers, primarily characterizing lesion load (number and

volume of lesions), or brain volume atrophy (which is accelerated in MS brains). However, volume loss associated with underlying neurodegeneration in MS over a 6-12 month period, while abnormal, may still be less than the precision of the volumetric measurement, such that volume changes in individual MS patients may only

be confidently diagnosed over 18-24 months and based on multiple measurements. The volumetric measurement also is most valid and precise only when obtained on the same MRI scanner each time, which is often impractical in the current US healthcare environment. Our approach is to derive quantitative MRI biomarkers from diffusion MRI (dMRI), which describe properties of

the microscopic environment and has been shown to outperform volumetrics. There has been a global effort to reparametrize the WM dMRI signal, through the standard model (SMI) of white matter, as the summation of tissue compartments typically over intra-axonal, extra-axonal and free-water compartments, enabling

disentangling inflammatory and degenerative processes. Aim 1. Curate database of age- and sex-matched normative reference SMI parameters. In collaboration with NYU’s Multiple Sclerosis Comprehensive Center (MSCC), we will retrospectively reprocess multi-shell dMRI data in adult populations for normal (N=544) and MS (N=418) patients to define clinically relevant SMI ranges.

Aim 2. Develop a software platform to automate biomarker analysis. Using data from Aim 1, an FDA- compliant backend software pipeline performing image processing, statistical comparisons and end-user reporting will be developed in Python and containerized for cross-platform generalization. The end-product is the

ms-MICSI report to aid clinicians in prognostic decision making and inform pharmaceutical drug trials. Aim 3. Verification and Validation of the ms-MICSI report will be performed on (A) 15 healthy volunteers, (B) 40 MS patients, 15 of which will return for a follow-up in 9 months to be imaged, and (C) scan-rescan (N=20

age/sex matched healthy volunteers) to evaluate bias and precision of SMI parameters. Cross sectional and longitudinal changes observed by ms-MICSI will be compared against commercially available volumetrics.