TABLE 2.
Classification performance of the YOLOv11 m model on the test set.
| Class | Precision % | Sensitivity % | Specificity % | F1 score % | Accuracy % | AP % | AUC |
|---|---|---|---|---|---|---|---|
| Ileocecal part | 99.97 | 86.21 | 99.96 | 92.59 | 97.15 | 98.13 | 99.39 [0.93,0.99] |
| Instrument | 94.07 | 98.23 | 95.83 | 96.1 | 96.8 | 99.55 | 99.69 [0.94,0.99] |
| Normal | 97.35 | 99.88 | 98.25 | 98.65 | 98.93 | 99.98 | 99.99 [0.96,1.00] |
| Overall (weighted average) | 96.58 | 96.44 | 97.64 | 96.38 | 96.44 | 99.42 | 99.75 [0.98,0.99] |
Weighted average metrics take into account the number of samples in each class, assigning greater weight to classes with larger sample sizes.