CVD prediction on Clinical Multimodal Dataset Extended (test)
93.2AUROCMedM2T
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MedM2TAblation Configuration=Full Model2025.10 | 93.2 | 67 | |
| MedM2TAblation Configuration=w/o Shared Encoder2025.10 | 93 | 66.6 | |
| MedM2TAblation Configuration=w/o Bi-Modal Attention2025.10 | 92.5 | 66.5 | |
| MultiBenchFusion Strategy=LF, Encoder Type=Our Encoder2025.10 | 92.3 | 66.1 | |
| HAIM2025.10 | 92.3 | 68 | |
| MedM2TAblation Configuration=w/o Pre-trained Encoder2025.10 | 91.7 | 65.3 | |
| MultiBenchFusion Strategy=LRTF, Encoder Type=Our Encoder2025.10 | 91.7 | 64.3 | |
| MultiBenchFusion Strategy=LRTF, Encoder Type=Original2025.10 | 91.6 | 64.9 | |
| MultiBenchFusion Strategy=LF, Encoder Type=Original2025.10 | 91.5 | 64.9 | |
| MultiModNEncoder Type=Our Encoder2025.10 | 91.1 | 63.1 | |
| MultiModNEncoder Type=Original2025.10 | 88.9 | 59.3 |