Table 4.

Top 10 combined features selected by the sparse linear regression with stability selection in the LOOCV experiments.

Feature	Frequency (%)	Feature	Frequency (%)
MCInc vs. MCIc_mixed		MCInc vs. MCIc_m12
CT: IPC_L	100	CT: IPC_L	100
CV: IPC_L	100	CT: cMFG_R	100
ND: MTG_L	100	CV: IPC_L	100
ND: PoCG_L	100	CV: MTG_L	100
ND: LING_R	100	ND: FP_L	100
NL: IPC_L	100	ND: MTG_L	100
CV: SMG_R	99	ND: PoCG_L	100
ND: PCC_R	97	ND: LING_R	100
CV: MTG_L	96	CV: SMG_R	87
ND: IPC_L	47	CT: BSTS_L	63
MCInc vs. MCIc_m6		MCInc vs. MCIc_m18
CT: IPC_L	100	CT: IPC_L	100
CT: MTG_L	100	CT: MTG_L	100
CV: IPC_L	100	CT: cMFG_R	100
ND: MTG_L	100	CT: PCUN_L	99
ND: LING_R	100	ND: PCC_R	99
CV: MTG_L	99	CT: IstCC_R	98
ND: PoCG_L	99	CT: BSTS_L	90
NL: ENT_L	88	CV: IPC_L	90
CV: SMG_R	85	ND: MTG_L	70
NL: IPC_L	74	CV: MTG_L	53

CT, cortical thickness; CV, cortical volume; CS, cortical surface area; NL, nodal length path; ND, nodal degree, L, left hemisphere; R, right hemisphere.