Figure 1.
An example of automated volumetric assessment.
A, FLAIR image from a patient with the usual extent of T2 hyperintense lesions seen in patients with clinically isolated syndrome. B, T1-weighted image with hypointense T1 lesions (arrows, lesion volume = 0.3 cm3). It is apparent that the T1 lesions were included in calculation of the overall brain volume (red area). C, T1-seighted image segmented by SIENAX. The T1 hypointense lesions (arrows) were misclassified into grey matter.
Table 1.
Demographic, clinical and MRI characteristics of the sample.
Figure 2.
Disability and volumetric MRI parameters in patients with CIS and in those converting to CDMS.
Dashed lines delineate 95% confidence intervals. Statistically significant p-values are shown. CIS, clinically isolated syndrome; CDMS, clinically definite multiple sclerosis; EDSS, Expanded Disability Status Scale; Gd+, gadolinium positive; MSFC, Multiple Sclerosis Functional Composite.
Table 2.
Relative risk of conversion to CDMS predicted by decrease in CC area at 6 months of CIS.
Figure 3.
Cumulative risk of CDMS by number of volumetric MRI predictors.
The MRI predictors found to be statistically significant by the logistic model were tested. These comprised decrease in corpus callosum area at 6 months ≥1%, and baseline T2 lesion volume ≥5 cm3. Hazard ratios with 95% confidence intervals are shown. CDMS, clinically definite multiple sclerosis, HR, hazard ratio.
Figure 4.
Effect of relapses on disability change and volumetric MRI parameters at 2 years of CIS.
Patients were categorised by the number of relapses during the 2-year follow-up period (0: n = 125, 1: n = 30, 2: n = 35, 3: n = 16, 4+: n = 11). Least-squares regression lines (dashed) and statistically significant p-values are shown. EDSS, Expanded Disability Status Scale; Gd+, gadolinium positive; MSFC, Multiple Sclerosis Functional Composite; T2LV, T2 lesion volume.