Skip to main content
. Author manuscript; available in PMC: 2023 Jan 17.
Published in final edited form as: Neuroimage. 2022 Nov 7;264:119737. doi: 10.1016/j.neuroimage.2022.119737

Table 8.

We compute the statistical difference of the learned connectivity matrices for OASIS ICA when predicting dementia, age and gender. The results show that the learned connectivity matrices are highly statistically different and SMN gets higher connectivity scores than DMN for dementia prediction whereas the opposite is seen for gender prediction.

Network 1 Network 2 Test Type P-value Avg. Connectivity 1 Avg. Connectivity 2

Dementia_All Age_All t-test 5e-22 0.323 0.168
manwhitneyu 1e-38
Dementia_All Gender_All t-test 2e-3 0.323 0.311
manwhitneyu 8e-4
Age Gender t-test 2e-301 0.168 0.311
manwhitneyu 1e-301
Dementia_DM Dementia_SM t-test 1e-7 0.478 0.645
manwhitneyu 8e-8
Age_DM Age_SM t-test 6e-1 0.294 0.308
manwhitneyu 6e-2
Gender_DM Gender_SM t-test 4e-1 0.527 0.555
manwhitneyu 1e-1
FNC_DM FNC_SM t-test 3e-2 0.487 0.580
manwhitneyu 7e-3
Dementia_DM Age_DM t-test 9e-6 0.478 0.294
manwhitneyu 5e-7
Dementia_DM Gender_DM t-test 2e-1 0.478 0.527
manwhitneyu 1e-1
Age_DM Gender_DM t-test 3e-7 0.294 0.527
manwhitneyu 5e-8
Dementia_SM Age_SM t-test 8e-34 0.645 0.308
manwhitneyu 3e-23
Dementia_SM Gender_SM t-test 4e-4 0.645 0.555
manwhitneyu 1e-4
Age_SM Gender_SM t-test 1e-18 0.308 0.555
manwhitneyu 4e-17