To apply Tact Based Spatial Statistics (TBSS)¹⁾　to evaluate the expansion of white matter damage in neuromyelitis optica (NMO) assessed by neurite orientation dispersion and density imaging (NODDI), as well as its relationship with disease severity.Twenty-four clinically confirmed NMO patients (20 women and 4 men; age 49.5 ± 11.9 years) and 23 age- and sex- matched healthy controls (control; 15 women and 8 men; age 45.9 ± 23.3 years) participated in this retrospective study. Informed consent was obtained from all the participants before evaluation. Diffusion data were acquired on a clinical 3T-MRI scanner (Philips Medical systems) as follows: TR/TE 4577/80 ms; 40 slices with thickness 3 mm; resolution 3×3 mm; imaging time of approximately 11 min; 3 b values (0, 1000, 2000) with diffusion encoding in 32 noncolinear directions for every b value. We generated FA maps by applying diffusion-weighted images with values of 0, 1000 sec/mm2. NODDI images including intracellular volume fraction (ICVF) and orientation dispersion index (ODI) map were computed by applying diffusion-weighted images with values of 0, 1000 and 2000 sec/mm2. TBSS was applied for voxel-wise analysis of FA, ODI, and ICVF maps. Group differences between NMO patients and controls were performed. Correlations with EDSS score in NMO patients group were also reported (adjusted for disease duration). The resulting statistical maps threshold was set at p < 0.05, with correction for multiple comparisons by using the threshold-free cluster enhancement (TFCE)¹⁾. All the anatomic information was based on the Johns Hopkins University white matter tractography atlas and the International Consortium for Brain Mapping (JHU-ICBM) DTI-81 white matter labels. In addition, we performed ROI study in the regions where significant correlation between FA and EDSS score, and between ICVF and EDSS score, were found in the TBSS study. ROIs were drawn by aligning common parts of the FA map and the registered ODI and ICVF maps based on the mean FA skeleton and by applying the JHU-ICBM DTI-81 white matter labels.In the voxel-wise group comparison, almost all the white matter tracts showed significantly reduced FA and increased ODI in NMO (Fig.1), while ICVF showed no significant differences. When we reset the threshold to P < 0.01, we found reduced FA in genu, body, and splenium of the corpus callosum, right and left anterior corona radiate, right and left superior corona radiate, right and left posterior corona radiate, right posterior thalamic radiation (including optic radiation), and right anterior limb of internal capsule. Increased ODI was also found in genu and body of the corpus callosum, left superior corona radiat, and left posterior corona radiate. In the correlation analyses, significant inverse correlations were observed between FA and EDSS score in genu and body of corpus callosum, left anterior limb of internal capsule, and left and right anterior corona radiate (Fig.2). ICVF also inversely correlated with EDSS score in the same significant white matter tracts of FA, in addition to right anterior limb of internal capsule, left posterior limb of internal capsule, and left external capsule (Fig.2). ODI showed no significant correlation. ROI analysis for significantly correlated area with FA (5 areas) and ICVF (8 areas) showed that only ICVF of genu of corpus callosum were significant between relatively severe patients and mild patients (severe vs. mild, 0.583 vs 0.647; P = 0.0008, significance level was set at 0.05/13 = 0.0038, applying Bonferroni correction, Mann Whitney U test).A recent study²⁾ in NMO patients showed extensive white matter damage, which could be related with not only Wallerian degeneration resulting from lesions of spinal cord or optic tracts but also demyelination by using diffusion-tensor (DT) MRI imaging. Similarly, we found extensively reduced FA of white matter tracts in NMO patients. Our results also demonstrated that extensively increased ODI, which suggests a loss of fiber coherence³⁾, was related with white matter damage in NMO. In correlation analyses, FA and ICVF were inversely correlated with EDSS score. Moreover, we found that significant areas were observed more extensively in ICVF than FA, and ICVF of genu of corpus callosum was the only significant hallmark in relatively severe NMO patients. Our results suggests that compared to FA or ODI, loss of neurite density in genu of corpus callosum, which is estimated by ICVF³⁾, could be a more sensitive marker for disease severity.NMO patients showed extensively increased ODI compared with healthy controls. Moreover, ICVF of genu of corpus callosum could be a sensitive marker for disease severity in NMO patients.