Table 7.

Performance metrics for the third stage of the full classification problem using the original images (101-layer residual network).

Classification issues	Acc	Sensitivity	Specificity
1	0.8488 ± 0.0133	0.8769 ± 0.0534	0.7611 ± 0.0815
2	0.9768 ± 0.0202	0.8763 ± 0.0861	0.9964 ± 0.0096
3	0.7993 ± 0.0436	0.9739 ± 0.0345	0.6414 ± 0.0965
4	0.9033 ± 0.0191	0.9673 ± 0.0238	0.5908 ± 0.1175
5	0.9147 ± 0.0151	0.9436 ± 0.0103	0.5000 ± 0.1023
6	0.9893 ± 0.0106	0.9851 ± 0.0178	0.9903 ± 0.6605
7	0.9796 ± 0.0126	—	—
8	0.9148 ± 0.0196	0.9192 ± 0.0614	0.9152 ± 0.0343
9	0.9085 ± 0.0392	0.9286 ± 0.0633	0.8554 ± 0.0997
10	0.9154 ± 0.0963	—	—