PEYOLO: a wrist fracture detection network based on multi-level receptive field feature extraction and cross-scale fusion
-
By
-
Shuwei Zhang
-
Rong Tang
-
Jiong Mu
-
Shaohai Ren
-
May 11, 2026
-
Clinical Scorecard: PEYOLO: A Novel Network for Detecting Wrist Fractures Utilizing Multi-Level Feature Extraction and Cross-Scale Fusion Techniques
At a Glance
| Category | Detail |
|---|
| Condition | Wrist Fractures |
| Key Mechanisms | Multi-level receptive field feature extraction and cross-scale feature fusion |
| Target Population | Children and the elderly |
| Care Setting | Clinical practice, particularly in emergency settings |
Key Highlights
- PEYOLO improves mean Average Precision (mAP) by 1.4% over baseline models.
- Utilizes Parallel Dilated Multi-head Attention Module (PDMAM) for feature extraction.
- Introduces Efficient Multi-Scale Attention module for enhanced feature perception.
- Demonstrates high precision and inference speed for clinical diagnosis.
- Addresses challenges of traditional manual interpretation in radiology.
Guideline-Based Recommendations
Diagnosis
- Utilize X-ray imaging as the preferred method for diagnosing wrist fractures.
- Implement automated computer-aided diagnostic methods to enhance accuracy.
Management
- Adopt PEYOLO for rapid and accurate detection of wrist fractures in clinical workflows.
Monitoring & Follow-up
- Regularly evaluate the performance of PEYOLO against traditional methods.
Risks
- High workload and pressure on emergency physicians may lead to missed or misdiagnosed fractures.
Patient & Prescribing Data
Patients with suspected wrist fractures, particularly in emergency settings.
PEYOLO can significantly improve diagnostic accuracy and efficiency.
Clinical Best Practices
- Incorporate AI-based diagnostic tools like PEYOLO to support radiologists.
- Ensure continuous training and validation of AI models with diverse datasets.
Related Resources & Content
