Skip to main content
. 2014 Apr 18;9(4):e95540. doi: 10.1371/journal.pone.0095540

Table 4. The results of ROI specific ACM/FA disease score correlations and GLM analyses.

Region of interest Pearson's correlation GLM with correction for age and gender r2
ACM FA ACM FA ACM FA
R-superior longitudinal fasciculus −0.46(b*) NS 0.57(b*) 0.51(b*) 0.33 0.26
R-corticospinal tract −0.29(*) NS 0.50(b*) 0.50(b*) 0.25 0.25
R-Posterior corona radiate −0.62(b*) NS 0.70(b*) 0.53(b*) 0.49 0.28
L-Posterior corona radiate NS NS 0.52(b*) 0.53(*) 0.27 0.28

Legend: The first column lists of a WM region of interest delineated in the JHU-ICBM-DTI-81 atlas with the letters R and L indicating a specific hemisphere. The second and third columns show Pearson's correlations between the summary statistics of median ACM or median FA and the EDSS disease scores. Boneferroni correction is used to account for multiple ROI comparisons (n = 10) and the (b*) indicates a Boneferroni corrected significant correlation at a significance level of 0.05 and (*) indicates uncorrected significance (0.05 level) while NS means not significant. The fourth and fifth column gives the linear estimate of disease score prediction based on a general linear model (GLM) including covariates of age, gender and either the median ACM or median FA while r2 is the coefficient of determination, summarizing the percentage of variance explained by the model.