Table 3.
Healthy versus stage 1 | Accuracy | Sensitivity | Specificity | AUC |
Axial map | 0.6667 (0.5667 to 0.7667) |
1.0 (1.0 to 1.0) |
0.0 (0.0 to 0.0) |
0.5 (0.5 to 0.5) |
Thickness map | 0.8667 (0.8 to 0.9333) |
0.875 (0.7857 to 0.9524) |
0.8500 (0.7143 to 0.9630) |
0.8625 (0.7792 to 0.9375) |
Front elevation map | 0.8000 (0.7167 to 0.8833) |
0.8750 (0.7838 to 0.9535) |
0.6500 (0.4583 to 0.8235) |
0.7625 (0.6584 to 0.8562) |
Back elevation map | 0.8500 (0.7667 to 0.9167) |
0.8750 (0.7838 to 0.9524) |
0.8 (0.6471 to 0.9412) |
0.8375 (0.7457 to 0.9167) |
Concatenated | 0.9 (0.8333 to 0.9500) |
0.9245 (0.8537 to 0.9783) |
0.85 (0.7143 to 0.96) |
0.8875 (0.8103 to 0.9556) |
Majority voting | 0.9167 (0.8500 to 0.9667) |
0.9500 (0.8864 to 1.0) |
0.8500 (0.7143 to 0.9600) |
0.9000 (0.8242 to 0.9653) |
95% CI in brackets.
AUC, area under the curve.