Skip to main content
. 2020 Nov 19;42(4):1206–1222. doi: 10.1002/hbm.25287

TABLE 4.

Brain regions showing modular reorganization after permutational‐based analysis against a random model, at a cut‐off threshold of p < .01

Absolute agreement difference (+/−) Coordinate in MNI space Harvard‐Oxford cortical and subcortical structural atlas Network assignment a p‐Value b 1,000 permutations
23.28 (10.27, −13.01) (44, −53, 47) Angular gyrus, R (40%) FP task control <.001
26.25 (12.84, −13.39) (32, 14, 56) Middle frontal gyrus, R (42%) FP task control <.001 c , d
23.35 (10.14, −13.21) (55, −44, 37) Supramarginal gyrus, R (43%) Salience <.001
21.45 (10.59, −10.87) (48, 25, 27) Middle frontal gyrus, R (39%) FP task control .002
24.31 (10.56, −13.76) (39, 18, 39) Middle frontal gyrus, R (50%) FP task control .004 b
17.15 (8.14, −9.01) (−60, −25, 14) Parietal operculum cortex, L (35%) Auditory .004
16.46 (6.97, −9.48) (−5, −18, 34) Cingulate gyrus, L (29%) CO task control .006 c
16.35 (7.10, −9.25) (−34, 3, 4) Insular cortex, L (4%) CO task control .007 b
19.42 (8.28, −11.15) (47, −50, 29) Angular gyrus, R (56%) Default mode .009

Note: Absolute agreement difference (average of absolute value per node) and its decomposition into positive and negative contributory factors are listed on the first column. Nodes are labeled with the probabilistic Harvard‐Oxford cortical and subcortical structural atlas, using peak coordinate for each ROI.

Abbreviations: CO, cingulo‐opercular; DMN, default mode network; FP, frontoparietal; HOA, hip OA; KOA, knee OA; MNI, Montreal Neurological Institute; OA, osteoarthritis.

a

Network assignment in accordance with Figure 5.

b

p‐Values are one‐sided and calculated after randomly permutating participants over 1,000 iterations and generating a null model for reorganization estimates (agreement difference matrix).

c

Nodal differences were further validated in the KOA holdout sample.

d

HOA group at p < .05 (Table S1).