LAUSR

Background: White matter hyperintensity (WMH), also known as leukoaraiosis, is commonly visualized as abnormal T2 signal in the deep and subcortical white matter on Magnetic Resonance Imaging (MRI). It is also commonly associated with aging, diabetes, hypertension and cerebrovascular disease. The Fazekas (F) scoring system is a subjective tool commonly used to assess WMH, but no volumetric analysis has been published showing how the scores correspond to true quantities of white matter disease. Methods: MRIs performed on inpatients and outpatients at our tertiary care institution between 2015 and 2017 were reviewed and their relative WMH was scored by one author trained in using the Fazekas scale. Using 3D Slicer 4.9, manual segmentations of WMH were completed and a 3D model was created to quantify the amount of WMH. Univariate analysis and ANOVA tests were run to determine the association of each Fazekas score with volume of WMH. Results: Among the 198 patients in our study (53% female), 163 had WMH (F1 n=66, F2 n=49, F3 n=48). Ranges of WMH in each group were 0.1-8.3 mL in Fazekas 1 (mean = 3.7, SD = 2.3), 6.0-17.7 mL in Fazekas 2 (mean = 10.8, SD = 3.1), and 14.2-77.2 mL in Fazekas 3 (mean = 35.2, SD = 17.9); if 11 outliers above 50 mL were excluded, the range for Fazekas 3 was 14.2-47.0 mL (mean = 27.1, SD = 8.9). When comparing data between groups, both the comparison between F1+2 (t-value = 14.1, p<0.001) and F2+F3 (t-value = 9.62, p<0.001) were significant. Moreover, when comparing between the three groups, each range of values was found to be significant from one another (F = 151.3, p<0.001). Conclusion: When accurately trained in assigning Fazekas scores to patient’s WMH, each of the scores appears to represent an approximate range of distinct volumes for WMH. Studies have shown that the presence and extent of WMH is a predictor for future development of stroke. These results should be validated in subsequent studies.