AD diagnosis on ADNI (test)
83.6BaccLNE
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| LNEEvaluation Protocol=Fine-tune2021.03 | 83.6 | — | — | — | — | — | |
| SimCLREvaluation Protocol=Fine-tune2021.03 | 82.4 | — | — | — | — | — | |
| LSSLEvaluation Protocol=Fine-tune2021.03 | 82.1 | — | — | — | — | — | |
| LNEEvaluation Protocol=Frozen2021.03 | 81.9 | — | — | — | — | — | |
| MRI Upper BoundInput=Real MRI2026.03 | 81 | — | — | 85 | 84 | 77 | |
| AEEvaluation Protocol=Fine-tune2021.03 | 80.7 | — | — | — | — | — | |
| SiTInterpretable=true2025.06 | 80.5 | 80.5 | — | — | — | — | |
| X-SiTInterpretable=true2025.06 | 80.2 | 80.6 | — | — | — | — | |
| No pretrainEvaluation Protocol=Fine-tune2021.03 | 79.4 | — | — | — | — | — | |
| VAEEvaluation Protocol=Fine-tune2021.03 | 77 | — | — | — | — | — | |
| LSSLEvaluation Protocol=Frozen2021.03 | 74.2 | — | — | — | — | — | |
| EHR+MRI Upper BoundInput=Real EHR + Real MRI2026.03 | 73 | — | — | 62 | 71 | 75 | |
| SimCLREvaluation Protocol=Frozen2021.03 | 72.9 | — | — | — | — | — | |
| AEEvaluation Protocol=Frozen2021.03 | 72.2 | — | — | — | — | — | |
| SpiralNet++Interpretable=false2025.06 | 70.3 | 69 | — | — | — | — | |
| MIRAGEClassifier=MLP, Input=EHR + Syn MRI2026.03 | 70 | — | — | 75 | 63 | 78 | |
| SiT-NeTInterpretable=false, Adapted for cortical surface=true2025.06 | 68.5 | 68.6 | — | — | — | — | |
| MIRAGEClassifier=RF, Input=EHR + Syn MRI2026.03 | 68 | — | — | 78 | 55 | 81 | |
| Random Forest (RF)Input=EHR-only2026.03 | 67 | — | — | 80 | 39 | 86 | |
| MVAEInput=EHR2026.03 | 67 | — | — | 71 | 50 | 84 | |
| MIRAGEClassifier=LR, Input=EHR + Syn MRI2026.03 | 67 | — | — | 75 | 56 | 78 | |
| VAEEvaluation Protocol=Frozen2021.03 | 66.7 | — | — | — | — | — | |
| Logistic Regression (LR)Input=EHR-only2026.03 | 66 | — | — | 75 | 58 | 75 | |
| Conditional DiffusionInput=EHR2026.03 | 66 | — | — | 67 | 47 | 84 | |
| EHR → MRIInput=EHR2026.03 | 65 | — | — | 72 | 52 | 80 | |
| EHR-latent-decoderInput=EHR2026.03 | 64 | — | — | 72 | 48 | 81 | |
| Multi-Layer Perceptron (MLP)Input=EHR-only2026.03 | 62 | — | — | 72 | 42 | 82 | |
| S-UNetInterpretable=false, Encoder only=true2025.06 | 57.6 | 57.2 | — | — | — | — | |
| 3D ResNetBackbone=ResNet502026.01 | — | — | 89.17 | 92.02 | — | — | |
| AXIAL2026.01 | — | — | 85.92 | 84.73 | — | — | |
| M3T2026.01 | — | — | 86.03 | 88.83 | — | — | |
| Max (DINOv3)Backbone=ViT-B/16, Compression strategy=Max pooling2026.01 | — | — | 75.45 | 55.68 | — | — | |
| Max (I-JEPA)Backbone=ViT-H/14, Compression strategy=Max pooling2026.01 | — | — | 77.27 | 73.63 | — | — | |
| Mean (DINOv3)Backbone=ViT-B/16, Compression strategy=Mean pooling2026.01 | — | — | 77.13 | 65.77 | — | — | |
| Mean (I-JEPA)Backbone=ViT-H/14, Compression strategy=Mean pooling2026.01 | — | — | 85.02 | 90.29 | — | — | |
| MERLINMode=image encoder only2026.01 | — | — | 90.77 | 93.42 | — | — | |
| MVSC (DINOv3)Backbone=ViT-B/16, Compression strategy=MVSC2026.01 | — | — | 91.89 | 96.08 | — | — | |
| MVSC (I-JEPA)Backbone=ViT-H/14, Compression strategy=MVSC2026.01 | — | — | 93.35 | 97.16 | — | — | |
| PPAL2026.01 | — | — | 86.08 | 79.95 | — | — | |
| RAPTOR2026.01 | — | — | 89.86 | 92.95 | — | — | |
| ViTBackbone=ViT-S/162026.01 | — | — | 86.49 | 90.92 | — | — |