Disease Classification on NIH-CXR
84.3AUROCX-WIN
Evaluation results
| Method | Links | |
|---|---|---|
| X-WINPretrain. data=372K CXRs + 32K CT volumes, Backbone=ViT-Large, Evaluation protocol=linear probing2025.11 | 84.3 | |
| X-WINPretrain. data=372K CXRs + 32K CT volumes, Backbone=ViT-Base, Evaluation protocol=linear probing2025.11 | 84.1 | |
| CheXWorldPretrain. data=448K CXRs, Backbone=ViT-Base, Evaluation protocol=linear probing2025.11 | 83.3 | |
| Ark+Pretrain. data=704K CXRs, Backbone=Swin-Base, Evaluation protocol=linear probing2025.11 | 83.1 | |
| CheXFoundPretrain. data=987K CXRs, Backbone=ViT-Large, Evaluation protocol=linear probing2025.11 | 82.9 | |
| MaCoPretrain. data=377K CXRs w/ reports, Backbone=ViT-Base, Evaluation protocol=linear probing2025.11 | 82.7 | |
| RAD-DINOPretrain. data=LVD-142M + 838K CXRs, Backbone=ViT-Base, Evaluation protocol=linear probing2025.11 | 82.1 | |
| CARZeroPretrain. data=377K CXRs w/ reports, Backbone=ViT-Base, Evaluation protocol=linear probing2025.11 | 81.3 | |
| CXR-AlignPretrain. data=325K CXRs w/ reports, Backbone=ViT-Base, Evaluation protocol=linear probing2025.11 | 81 | |
| CheXAgentPretrain. data=1.07M CXRs w/ text, Backbone=ViT-Large, Evaluation protocol=linear probing2025.11 | 75.6 | |
| DINOv2Pretrain. data=LVD-142M, Backbone=ViT-Base, Evaluation protocol=linear probing2025.11 | 71.1 | |
| I-JEPAPretrain. data=ImageNet-1K, Backbone=ViT-Large, Evaluation protocol=linear probing2025.11 | 69.2 |