Conventional MRI imaging is highly sensitive for detection of MS lesions, which are characteristically hyperintense on a T2 weighted images, however this technique lacks pathological specificity and as a result edema, demyelination, remyelination, axonal loss and gliosis are indistinguishable (1). QSM describes the underlying total magnetic susceptibility sources in a voxel (2). In MS lesions, removal of diamagnetic degraded myelin debris and iron accumulation both increase susceptibility measured on QSM (3). QSM can help identify myelin and iron content changes during an MS lesion’s lifetime. The aim of this study was to combine myelin water fraction (MWF) imaging and QSM to investigate the sensitivity of QSM to detect myelin pathology in the tissue immediately surrounding MS lesions.

Imaging experiment: Eight MS patients scanned on 3T GE scanner (HDxt 16.0) using 8-channel phased array coil from February to March 2015 were shortlisted from our ongoing database. All patients had a Fast Acquisition with Spiral Trajectory and T2prep (FAST-T2), which provides high quality reproducible MWF and GRE images. QSM maps were reconstructed from a three dimensional T2* -weighted spoiled multi-echo GRE sequence by using morphology-enabled dipole inversion (MEDI) method (4). Scanning parameters: T2-FLAIR (1.2x0.6x0.6mm) and FAST-T2 (1x1x5mm).FAST-T2 parameters and QSM parameters have been previously published (4, 5)

Post Processing and Data analysis: MWF maps were obtained by analyzing FAST-T2 data with our Multi-Gaussian post-processing algorithm (5). MWF maps and T2-FLAIR was co-registered to QSM magnitude using FSL’s linear FLIRT transformation (6). Specific T2 and QSM lesions were defined on T2-FLAIR by trained neuro-radiologist. For each MS lesion, the ROI created manually in 3D on both T2w and QSM. Refer to abstract # 5706 for a presentation demonstrating that QSM volume exceeds that of T2. For this analysis, a new ROI was created to represent the difference between QSM and T2 signals (QSM ROI - T2 lesion ROI) within identified lesions. New ROI’s were designated as QSM_rim and T2_lesion. To reduce error related to noise we removed T2 lesions that were smaller than 100 voxels and failed registration and QSM_rim lesions that were smaller than 25 voxels.

We analyzed in total 8 subjects having 61 rim lesions. Out of these 61 lesions, 10 lesions were excluded to meet our size criteria of minimum lesion size of 100 voxels. Figure 1 shows an example of a single slice of QSM, T2-FLAIR, and MWF map (top row) before and after overlaying the lesion mask (bottom row). We can clearly see the segmented T2_lesion ROIs and QSM_rim maps are well co-registered with each imaging modality. In addition, it can be appreciated that a lesion on QSM exceeds the volume of a T2 lesion. The average volume of the 51 T2_lesion was 258.24 mm3 and QSM_rim was 112.90 mm3. Two additional lesions were further excluded given the QSM_rim ROI’s were smaller than 25 voxels. Table 1 shows mean (SD) values of QSM_rim and T2_lesion ROI’s of the remaining 49 lesions. The QSM of T2_lesions ROI’s (28.84) was significantly higher than the QSM_rim ROI (23.28), p<0.0001. The MWF of T2_lesion ROI’s (0.058) was significantly lower than QSM_rim ROI (0.087), p<0.0001. We present a multi-modality approach to the study of MS lesions. We successfully co-registered each modality to demonstrate changes in QSM and MWF beyond the boundaries of the conventional T2 lesion. The volume of QSM lesions exceeded that observed on conventional T2 imaging Through our multi-modality imaging approach, we were able to demonstrate that MWF, an indirect measure of myelin, was higher in the QSM_rim ROI, suggesting that less demyelination was present in the QSM rim, which is consistent with the observed decrease in susceptibility. Importantly, the final MWF value, on average, was much lower than our previously reported healthy control lesion-based white matter MWF value (0.16) (7). In conclusion, the QSM volume, which exceeds the T2 lesion volume, represents myelin damage and thus is more sensitive to pathological tissue damage than conventional T2 imaging. More importantly, given that iron changes also influence susceptibility, QSM represents a combined measure of disease activity. We plan to further explore the use of myelin water imaging to extract the contribution of iron and myelin to the susceptibility signal change, which would then further advance QSM as a biomarker for MS.