OCT
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Anomaly Classification | OCT 17 | AUC99.93 | 54 | |
| Retinal Anomaly Detection | OCT 2017 (test) | F1 Score99.3 | 28 | |
| Retinal Disease Classification | OCT 2017 (test) | Accuracy99.8 | 24 | |
| Medical Image Segmentation | OCT | Dice (%)50.93 | 18 | |
| Top-k localization precision and sensitivity | OCT | Top-k Precision86 | 14 | |
| Model Explainability Faithfulness | OCT | AUDC85.2 | 14 | |
| Retinal Disease Classification | OCT-C8 (test) | Overall Accuracy (OA)95.25 | 13 | |
| Medical Diagnosis Classification | OCT | F1 Score (%)96.3 | 12 | |
| Anomaly Detection | OCT 2017 | Image-level AU-ROC99 | 12 | |
| Semantic Segmentation | OCT (test) | Relative Performance95.4 | 11 | |
| Anomaly Detection | OCT 2017 (test) | I-AUROC99.61 | 9 | |
| OCT to OCTA translation | OCT OCTA Projection Map 2024 | MAE0.0087 | 8 | |
| OCT Reconstruction | OCT 15 volumes (test) | PSNR28.99 | 7 | |
| Image-level Anomaly Detection | OCT 17 | AUC0.821 | 7 | |
| Medical Image Classification | OCT | ResNet Accuracy96.34 | 7 | |
| Image Segmentation | Healthy OCT (train) | Dice Score (%)91.9 | 6 | |
| Image Reconstruction | Healthy OCT (train) | MAE5.24 | 6 | |
| Anomaly Detection | OCT | Image-level AUC99.7 | 3 | |
| Generative Modeling | OCT Dataset (five-fold cross-validation) | KID49.1 | 3 | |
| Anomaly Classification | OCT 17 (test) | AUC95.4 | 3 | |
| Medical Image Classification | OCT-11k Target domain (Topcon) | Accuracy79.78 | 3 | |
| Medical Image Classification | OCT-11k (Source domain (Zeiss)) | Accuracy85.33 | 3 | |
| Image Segmentation | Healthy OCT (test) | Dice Score86.5 | 2 | |
| Image Reconstruction | Healthy OCT (test) | MAE (%)6.04 | 2 |
Showing 24 of 24 rows