Table 2 The DSC and HD values of all organs at risk (n = 9) of the experts’ manual contours and autocontours, listed from the best to the lowest performance
From: Evaluation of deep learning-based autosegmentation in breast cancer radiotherapy
Rank | DSC | HD |
|---|---|---|
Average (standard error) | ||
1 | 0.903 (0.022) | 5.142 (0.897)* |
2 | 0.896 (0.020)* | 5.327 (1.208) |
3 | 0.887 (0.021) | 5.477 (1.018) |
4 | 0.886 (0.020) | 5.615 (1.181) |
5 | 0.882 (0.022) | 5.780 (1.240) |
6 | 0.881 (0.020) | 6.431 (1.606) |
7 | 0.881 (0.025) | 6.447 (1.979) |
8 | 0.880 (0.030) | 6.461 (1.340) |
9 | 0.877 (0.022) | 6.501 (1.357) |
10 | 0.874 (0.027) | 6.724 (2.099) |
11 | 0.870 (0.029) | 7.636 (2.338) |