Multi-scale feature refinement network for lower limb fracture detection in X-ray images
-
By
-
Zhengguo Wan
-
Yanling Wang
-
Rong Tang
-
Ke Zhang
-
Penghua Liu
-
Zheyu Zhao
-
Yujie Shi
-
Huaran Huo
-
July 2, 2026
-
Clinical Scorecard: Multi-Scale Feature Enhancement Network for Detecting Lower Limb Fractures in X-Ray Imaging
At a Glance
| Category | Detail |
|---|
| Condition | Lower limb fractures |
| Key Mechanisms | Multi-scale feature refinement through Adaptive Feature Perception Block and Multi-Scale Dilated Attention Module |
| Target Population | Patients with lower limb fractures, particularly in orthopedic emergency settings |
| Care Setting | Orthopedic emergency departments |
Key Highlights
- MFRNet achieves 89.1% mAP and 52.2% mAP50-95 with only 3.8 M parameters
- Addresses challenges of varying lesion scales and ambiguous fracture boundaries
- Utilizes deep learning-based object detection algorithms for improved accuracy
- Proposes innovative modules to enhance feature discriminability
- Demonstrates significant performance improvement over existing models
Guideline-Based Recommendations
Diagnosis
- Utilize plain X-ray radiography as the first-line imaging modality for lower limb fractures
Management
- Implement computer-aided detection systems to assist in rapid localization of fractures
Monitoring & Follow-up
- Monitor for delayed or inaccurate diagnosis to prevent complications such as malunion
Risks
- Inaccurate detection can lead to malunion, delayed union, or nonunion of fractures
Patient & Prescribing Data
Individuals with suspected lower limb fractures
Rapid and precise detection is critical for effective treatment planning
Clinical Best Practices
- Incorporate advanced imaging techniques to improve fracture detection accuracy
- Utilize multi-scale feature extraction methods in diagnostic imaging
Related Resources & Content
