Research map / Detection and segmentation / CNN features
FracSegmentator: Fracture Instance Segmentation with Trauma-Prior-Guided Contrastive Learning
Yanzhen Liu, Sutuke Yibulayimu, Yang Zhou, Yudi Sang, Yu Wang
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 |
|---|---|---|---|
| add trauma prior memory attention (TPMA) (ablation) | FracSegmentator · Merged validation set | DSC: 94.32 -> 97.37 | Improved |
| add contrastive loss (ablation) | FracSegmentator · Merged validation set | DSC: 96.48 -> 97.37 | Improved |
| add feature calibration attention decoder (ablation) | FracSegmentator · Merged validation set | DSC: 96.61 -> 97.37 | Improved |
| add TPMA and contrastive learning together (ablation) | FracSegmentator · Merged validation set | DSC: 93.06 -> 97.37 | Improved |
| use BICM three-class mask instead of merging boundary and interior (ablation) | FracSegmentator · Merged validation set | DSC: 90.98 -> 97.37 | Improved |
| use BICM instead of merging boundary and contact (ablation) | FracSegmentator · Merged validation set | DSC: 90.73 -> 97.37 | Improved |
| CFS dilation radius 5mm instead of 1mm (ablation) | FracSegmentator · Merged validation set | DSC: 96.58 -> 97.37 | Improved |
| CFS dilation radius 5mm instead of 7mm (ablation) | FracSegmentator · Merged validation set | DSC: 97.20 -> 97.37 | Improved |
| FracSegmentator vs MedNext | FracSegmentator · SacrumFrac | DSC: 94.59 -> 97.59 | Improved |
| FracSegmentator vs prior PENGWIN SOTA | FracSegmentator · PENGWIN | IoU: 92.96 -> 95.72 | Improved |
Closest papers by meaning
- Scalable Geometric Fracture Assembly via Co-creation Space among Assemblers (AAAI 2024 · similarity 0.80)
- BLS-GAN: A Deep Layer Separation Framework for Eliminating Bone Overlap in Conventional Radiographs (AAAI 2025 · similarity 0.80)
- Integrating Deep Metric Learning with Coreset for Active Learning in 3D Segmentation (NeurIPS 2024 · similarity 0.80)
- Segmenting Maxillofacial Structures in CBCT Volumes (CVPR 2025 · similarity 0.80)
- Domain Generalized Medical Landmark Detection via Robust Boundary-Aware Pre-Training (AAAI 2025 · similarity 0.80)
- Training Like a Medical Resident: Context-Prior Learning Toward Universal Medical Image Segmentation (CVPR 2024 · similarity 0.79)
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.
