Algorithm performances in detection of any pterygium
| Testing sets | Detection of any pterygium | ||||
| AUC % (95% CI) | Sensitivity % (95% CI) | Specificity % (95% CI) | Accuracy % (95% CI) | MCC | |
| Internal: | |||||
| SEED | 99.5 (99.0 to 99.9) | 98.6 (97.0 to 99.7) | 99.0 (97.7 to 100.0) | 98.8 (98.0 to 99.7) | 0.976 |
| External: | |||||
| Set 1 | 99.1 (97.4 to 99.9) | 95.9 (89.2 to 100.0) | 98.5 (97.8 to 98.9) | 98.4 (97.8 to 98.9) | 0.782 |
| Set 2 | 99.7 (99.4 to 99.8) | 100.0 (100.0 to 100.0) | 88.3 (87.2 to 89.4) | 88.6 (87.5 to 89.7) | 0.400 |
AUC, area under the curve; MCC, Matthew’s correlation coefficient; SEED, Singapore Epidemiology of Eye Diseases.