Table 3.
Comparison between pairs of annotators and all annotators across all datasets.
| Annotator | Sørensen-Dice coefficient | Mean volume difference (cm3) | Mean volume difference (%) |
|---|---|---|---|
| A vs. B | 0.900 ± 0.056 | −0.003 ± 0.030 | −0.8% ± 16.5 |
| A vs. C | 0.907 ± 0.048 | 0.006 ± 0.029 | 2.2% ± 14.7 |
| B vs. C | 0.902 ± 0.046 | 0.009 ± 0.030 | 3.0% ± 15.1 |
| Fleiss’ Kappa | RMSE (cm3) | RMSE (%) | |
| All | 0.903 ± 0.046 | 0.015 ± 0.010 | 6.9 ± 4.7 |
The metrics were calculated over the 3D tumor volume. RMSE was performed between the volume of each annotator and the mean volume across all annotators.