Analyzing brain structural differences among undergraduates with different grades of self-esteem using multiple anatomical brain network

Table 1 Classification performance using different feature types

	ACC (%)	AUC (%)	SEN (%)	SPE (%)	Y (%)	F (%)	BAC (%)
Network features in the 4th layer	90.69	96.63	87.72	90.65	86.74	77.38	88.77
Network features in the 3rd layer	88.31	84.27	85.32	84.29	88.33	82.62	85.74
Network features in the 2nd layer	89.59	76.65	78.53	75.94	73.24	69.18	67.77
Network features in all layers	92.59	91.93	91.51	90.91	93.27	87.18	91.49
ROI features in the 4th layer	88.69	85.63	87.72	87.65	86.74	77.38	88.77
ROI features and network features in the 4th layer	94.41	95.58	94.42	93.41	92.47	92.82	92.64
Multilevel (ROI features in the 4th layer and network features in all layers)	97.26	99.88	97.27	97.41	97.12	94.53	97.27

ACC: accuracy; AUC: area under receiver operating characteristic curve; SEN: sensitivity; SPE: specificity; Y: Youden’s index; F: F-score; BAC: balanced accuracy

ISSN: 1475-925X