Table 3.

Results of transfer learning on four datasets.

Dataset	Classification	Sensitivity	Specificity	Precision	EER	FPR	F1 Score	Accuracy	Learning Time
OASIS (GM)	Binary	0.89	0.84	0.84	0.13	0.16	-	0.8621	93 min 6 s
	Multiple	60.28%	49.20%	49.20%	0.40	0.51	54.18%	0.6028	76 min 14 s
OASIS (WM)	Binary	0.66	0.92	0.89	0.2	0.08	-	0.8046	120 min 4 s
	Multiple	37.01%	35.25%	35.25%	0.63	0.65	36.11%	0.3701	113 min 23 s
OASIS (CSF)	Binary	0.66	0.88	0.85	0.22	0.12	-	0.7816	118 min 13 s
	Multiple	67.33%	62.63%	62.63%	0.33	0.38	64.90	0.6733	110 min 24 s
OASIS (Un-segmented)	Binary	1	0.82	0.847	0.1	0.18	-	0.8966	125 min 20 s
	Multiple	92.85%	74.27%	74.27%	0.07	0.26	82.53	0.9285	117 min 15 s