Table 8.

Performance metrics for the third stage of the full classification problem using the original images (50-layer residual network).

Classification issues	Acc	Sensitivity	Specificity
1	0.6668 ± 0.1483	0.5920 ± 0.4134	0.7186 ± 0.3171
2	0.8500 ± 0.0374	1 ± 0	0.1468 ± 0.2541
3	0.8125 ± 0.0779	0.9119 ± 0.0825	0.6756 ± 0.2605
4	0.7282 ± 0.1640	0.26 ± 0.6	0.8298 ± 0.3029
5	0.6095 ± 0.3356	0.7 ± 0.4203	0.6028 ± 0.4396
6	0.7047 ± 0.2035	0.9851 ± 0.0181	0.9903 ± 0.6605
7	0.4721 ± 0.2132	—	—
8	0.7234 ± 0.1521	0.5393 ± 0.3103	0.9137 ± 0.0898
9	0.8137 ± 0.0778	0.9119 ± 0.0825	0.7728 ± 0.0989
10	0.8776 ± 0.0695	—	—