Vertebra Level Fracture Recognition on RSNA Dataset
0.9975AccuracyRasul et al.
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Rasul et al.Dataset Split=Subset (4200 images), Test Set Size=400 Images, Major Drawback=Dataset was balanced before splitting.2026.01 | 0.9975 | — | — | — | — | |
| Yaseen et al.Dataset Split=Subset (14434 slices), Test Set Size=2887 slices, Major Drawback=Dataset was balanced before splitting.2026.01 | 0.978 | — | — | 97.8 | — | |
| Kim et al.Dataset Split=Full, Test Set Size=404 Patients2026.01 | 0.949 | — | — | — | — | |
| Proposed Method (Score Fusion)Test Set Size=2019 Patients2026.01 | 0.9451 | — | — | 68.15 | — | |
| RSNA First Place WinnerDataset Split=Full, Test Set Size=2019 Patients, Cross Validation=true2026.01 | 0.9405 | — | — | 69.74 | — | |
| Proposed MethodTest Set Size=2019 Patients, Cross Validation=true2026.01 | 0.9354 | — | — | 66.54 | — | |
| Chlad et al.Dataset Split=Subset, Test Set Size=4330 slices, Major Drawback=Dataset was balanced before splitting.2026.01 | — | 98 | 98 | — | — | |
| Sutradhar et al.Dataset Split=Subset (235 Patients), Test Set Size=24 Patients, Major Drawback=Dataset was small2026.01 | — | — | — | — | 93 |