Article Figures & Data
Figures
- FIG 1.
In the proposed MAP, atrophy (hypertrophy) was estimated on the basis of CTh maps (A). The cortical surface was subparcelled into 1000 “patches” (black lines in B), and the average CTh for each parcel was calculated (B). By referencing the value of each parcel to an age-/sex-matched control group, P maps can be calculated (C) so that the lower end of the spectrum is suggestive of “atrophy” (cool colors) and the higher end of “hypertrophy” (hot colors). MAP uses the fraction of any “atrophic” patches, ie, the TPF, as a metric to estimate cortical disease burden.
- FIG 2.
We investigated whether MAP differentiates patients with MS from HC. We found that both the primary outcome of MAP, ie, the TPF (A), as well as its opposite, the thick patch fraction (B), significantly discriminate MS from HC. Notice the opposed directionality: MS yields more thin but fewer thick patches. Boxplots show median (notched), IQRs, and outliers (dots); boxplot width reflects square root of sample sizes.
- FIG 3.
We investigated whether MAP discriminates among MS clinical phenotypes: Especially, the TPF (A) yielded high potential for classifying among MS clinical phenotypes because it was significantly different between PMS and HC/CIS/RRMS and, moreover, discriminated RRMS from HC. The thick patch count (B) was less sensitive and only differentiated SPMS and RRMS from HC. Boxplots show median (notched), IQRs, and outliers (dots); boxplot width reflects the square root of sample sizes. The asterisk denotes statistical significance at P < .05; double asterisks, P < .001 in post hoc testing.
- FIG 4.
We tested associations of MAP with established scores of MS disease burden: We found that the primary outcome of MAP, ie, the TPF, was significantly associated with EDSS scores (A), the cognitive MuSIC score (B), the fatigue MuSIC score (C), and lesion volume (D).
- FIG 5.
MAP allows visualizing individual brain maps with regionally specific estimations of atrophy (blue) or hypertrophy (red) based on comparisons with a matched reference group (A). This method may be further refined anatomically; for example, instead of summing thin patches across the entire cortex, one can specify super-ROIs (B) such as motor (red), frontal (purple), parietal (yellow), insular (light blue), temporal (green), and visual (dark blue) cortices. Thin/thick patch fractions can then be calculated for each of these super-ROIs and used for an individual and regionally specific illustration, eg, in a spider plot (C).
Tables
Demographic details of the study population
MS Clinical Profiles All Patients with MS HC Subgroups T Test (W)a/χ2 Test (C2)b CIS RRMS PMS CamCAN (External) TUM (Internal) T Test (W)/χ2 Test (C2) between HC No. subjects 19 465 17 501 651 89 NA NA Age (mean) (yr) 58.73 (SD, 12.06) 40.14 (SD, 9.94) 55.58 (SD, 5.23) 40.56 (SD, 10.22) 54.26 (SD, 18.58) 37.36 (SD, 15.06) W: t (150.61) = 9.78, P < .001 W: t (102.87) = −1.92, P = .056 Sex (female/male) 13:6 305:160 10:7 328:173 330:321 59:30 C2: χ2 (1, N = 740) = 7.03, P = .008 C2: χ2 (1, N= 590) < 0.001, P = .976 Dominant hand (R/L) 19/0 415/50 16/1 450/51 NA NA NA NA EDSS (median) (IQR) 0.0 (1.0) 1.5 (2.0) 4.25 (1.75) 1.5 (2.0) NA NA NA NA Cognition MuSIC (mean) 28.05 (SD, 3.01) 27.10 (SD, 3.66) 25.82 (SD, 4.05) 27.09 (SD, 3.66) NA NA NA NA Fatigue MuSIC (mean) 6.21 (SD, 2.64) 7.58 (SD, 4.59) 11.65 (SD, 3.32) 7.67 (SD, 4.56) NA NA NA NA Lesion volume, (mean) (mL) 0.98 (SD, 1.23) 5.39 (SD, 8.41) 11.78 (SD, 9.01) 5.45 (SD, 8.40) NA NA NA NA
Jump to section
Related Articles
Cited By...
- No citing articles found.