Derm7pt
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Medical Classification | Derm7pt | Accuracy84.69 | 48 | |
| Skin lesion recognition | Derm7pt | mAP76.7 | 20 | |
| Dermoscopy Image Classification and Concept Prediction | Derm7pt asymmetric filtering (test) | Label Accuracy85 | 19 | |
| Concept Prediction | Derm7pt symmetric filtering (test) | Concept Accuracy70 | 19 | |
| Label Classification | Derm7pt symmetric filtering (test) | Accuracy90 | 19 | |
| Skin Lesion Classification | Derm7pt | Expected Calibration Error (ECE)2.2 | 15 | |
| Skin Lesion Classification | derm7pt clinical (out-of-distribution) | ROC AUC0.7396 | 12 | |
| Skin Lesion Classification | derm7pt dermoscopic (out-of-distribution) | ROC AUC0.8369 | 12 | |
| Diagnosis classification | Derm7pt (test) | F177.3 | 12 | |
| Concept Annotation | Derm7pt | F1-Macro65.06 | 11 | |
| Concept Annotation | Derm7pt (test) | AUROC69 | 11 | |
| Concept Prediction | Derm7pt | AUC92.1 | 10 | |
| Multimodal Skin Lesion Classification | Derm7pt (test) | Sensitivity (SEN)79.87 | 9 | |
| Skin disease diagnosis | Derm7pt 1.0 (test) | AUC89.48 | 9 | |
| Multimodal Diagnosis | Derm7pt | Parameters28.47 | 6 | |
| Skin Cancer Image Retrieval | Derm7pt (5-fold cross-val) | mAP (Fold 1)83.8 | 5 | |
| Retrieval | Derm7pt | Top-1 Accuracy79.3 | 5 | |
| Melanoma inference | Derm7pt (test) | BA (t=1)69.1 | 5 | |
| 7-point criteria classification | Derm7pt (test) | BA (PN)68.6 | 5 | |
| Medical Image Classification | Derm7pt (test) | Accuracy92.29 | 4 | |
| Concept Retrieval | Derm7pt | P@10 (Streaks)100 | 3 | |
| Skin Cancer Classification | Derm7pt | ROC AUC0.76 | 1 |
Showing 22 of 22 rows