Research map / Detection and segmentation / CNN features
Window Loss for Bone Fracture Detection and Localization in X-ray Images with Point-based Annotation
Xinyu Zhang, Yirui Wang, Chi-Tung Cheng, Le Lu, Adam P. Harrison, Jing Xiao, Chien-Hung Liao, Shun Miao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use Window Loss with point-based supervision instead of image-classification or box-detection baselines | Pelvic X-ray fracture detection · Pelvic X-ray fracture detection | AUROC: 0.974 -> 0.983 | Improved |
| use Window Loss with point-based supervision instead of the strongest baseline | Pelvic X-ray fracture detection · Pelvic X-ray fracture detection | FROC score: 88.1% -> 89.9% | Improved |
| use KLD rather than MSE as the Window Loss divergence | Window Loss · Pelvic X-ray fracture detection | AUROC: 0.979 -> 0.983 | Improved |
| use MSE rather than KLD as the Window Loss divergence | Window Loss · Pelvic X-ray fracture detection | FROC score: 89.6% -> 89.9% | Improved |
| increase rℓ from 50 to 100 (ablation) | WL-MSE · Pelvic X-ray fracture detection | FROC score: 89.9% -> 86.4% | Hurt |
| increase rℓ from 50 to 150 (ablation) | WL-MSE · Pelvic X-ray fracture detection | FROC score: 89.9% -> 82.0% | Hurt |
| reduce τ from 2 to 0.4 (ablation) | WL-MSE · Pelvic X-ray fracture detection | FROC score: 89.9% -> 89.3% | Hurt |
| increase τ from 2 to 5 (ablation) | WL-MSE · Pelvic X-ray fracture detection | FROC score: 89.9% -> 88.2% | Hurt |
| increase τ from 2 to 10 (ablation) | WL-MSE · Pelvic X-ray fracture detection | Recall@0.1 FP/image: 87.4% -> 69.6% | Hurt |
| set ru=rℓ=50 and τ=0, producing binary bounds (ablation) | WL-MSE · Pelvic X-ray fracture detection | FROC score: 89.9% -> 86.1% | Hurt |
| reduce ru from 200 to 50 (ablation) | WL-MSE · Pelvic X-ray fracture detection | FROC score: 89.9% -> 87.2% | Hurt |
| increase rℓ from 50 to 100 (ablation) | WL-KLD · Pelvic X-ray fracture detection | FROC score: 89.6% -> 85.4% | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- FCOS — FCOS: Fully Convolutional One-Stage Object Detection (ICCV 2019) · better on 3 of 3 shared results
- Faster-RCNN — Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (TPAMI 2016) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- OralXrays-9: Towards Hospital-Scale Panoramic X-ray Anomaly Detection via Personalized Multi-Object Query-Aware Mining (CVPR 2025 · similarity 0.83)
- RAM-W600: A Multi-Task Wrist Dataset and Benchmark for Rheumatoid Arthritis (NeurIPS 2025 · similarity 0.81)
- BLS-GAN: A Deep Layer Separation Framework for Eliminating Bone Overlap in Conventional Radiographs (AAAI 2025 · similarity 0.81)
- Mudslide: A Universal Nuclear Instance Segmentation Method (CVPR 2024 · similarity 0.81)
- Novel Class Discovery in Chest X-rays via Paired Images and Text (AAAI 2024 · similarity 0.81)
- Coupled Laplacian Eigenmaps for Locally-Aware 3D Rigid Point Cloud Matching (CVPR 2024 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
