Related Experiment Video
Updated: Mar 24, 2026
07:56
Digital Planimetry for Assessing Wound Closure Kinetics in a Mouse Model
Published on: January 10, 2025
1.4K
Scalable and objective wound infection screening from clinical images using deep learning.
Chao Wang1, Hongyu Wang2, Jianhong Hu3
1Department of Burn and Plastic Surgery, West China School of Medicine, Sichuan University, Sichuan University Affiliated Chengdu Second People's Hospital, Chengdu Second People's Hospital, Chengdu, China.
Frontiers in Public Health
|March 4, 2026
Summary
This study developed a deep learning framework for automated wound infection detection using clinical images. The AI tool shows potential for earlier, more consistent diagnosis, aiding antimicrobial stewardship.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Wound Care
Background:
- Wound infection is a significant complication delaying healing and increasing costs.
- Current screening lacks rapid, objective, and scalable methods, especially in resource-limited settings.
- Inappropriate antimicrobial use is a concern in wound management.
Purpose of the Study:
- To develop and evaluate a deep learning (DL) framework for automated wound infection detection.
- To improve diagnostic consistency and support public health-oriented wound management.
- To assess the performance of DL models using clinical wound images.
Main Methods:
- Trained and evaluated DL models, including Swin Transformer and CNNs, on 4,000 clinical wound images.
- Assessed model performance using accuracy, AUC, and F1-score.
- Compared model predictions with non-specialist clinician assessments for real-world applicability.
Main Results:
- The Swin Transformer model achieved high performance: 0.9025 accuracy, 0.9546 AUC, and 0.9042 F1-score.
- DL model predictions reduced diagnostic variability compared to non-specialist clinicians.
- The model enabled earlier and more consistent recognition of wound infections.
Conclusions:
- DL applied to wound images offers a scalable, objective approach for infection screening.
- These tools can support earlier detection and reduce diagnostic variability.
- Potential to improve wound management and antimicrobial stewardship, especially in public health settings.
