List your company

Patient-level Fracture Detection on RSNA

84.39AccuracyRSNA First Place Winner
46.128456.061765.99575.9283Jan 21, 2026

Evaluation results

MethodLinks
RSNA First Place WinnerDataset Split=Full, Test Set Size=2019 Patients, Cross Validation=true2026.0184.39———82.56—
Proposed Network (Score Fusion)Input=2.5D (Slice + Max Stacks), View=Axial, Aggregation Method=If Any2026.0183.0693.5469.3195.5479.5890.95
Proposed Network (Score Fusion)Input=2.5D (Slice + Max Stacks), View=Axial, Aggregation Method=Adaptive Average2026.0183.0682.0182.5283.5582.2683.04
Proposed Method (Score Fusion)Test Set Size=2019 Patients2026.0183.06———82.26—
Proposed NetworkInput=2.5D (Slice Stacks), View=Axial, Aggregation Method=If Any2026.0182.6786.1576.1880.580.6689.09
Proposed MethodTest Set Size=2019 Patients, Cross Validation=true2026.0182.67———80.66—
Proposed NetworkInput=2.5D (Max Projection Stacks), View=Axial, Aggregation Method=If Any2026.0181.1888.8769.7977.0777.8789.15
Mahajan et al.Dataset Split=Full2026.0176—————
DenseNet121Input=2D (whole), View=Coronal + Sagittal, Aggregation Method=Score Fusion2026.0166.6766.8659.5973.1163.0169.76
EfficientNetB0Input=3D, Aggregation Method=If Any2026.016274.1430.5690.4341.3865.79
DenseNet121Input=2D, View=Axial, Aggregation Method=If Any2026.0161.4265.6439.7581.149.5134.04
DenseNet121Input=2D (whole), View=Coronal, Aggregation Method=None2026.0160.9959.1658.5563.2158.8564.63
DenseNet121Input=2D (whole), View=Sagittal, Aggregation Method=None2026.0159.2664.5832.1283.9642.959.26
DenseNet121Input=2D, View=Sagittal, Aggregation Method=If Any2026.0148.3447.9398.652.6564.5144.94
DenseNet121Input=2D, View=Coronal, Aggregation Method=If Any2026.0147.647.6100064.546.22