Automated chronic wounds medical assessment and tracking framework based on deep learning.

Area of Science:

• Medical Imaging
• Artificial Intelligence in Healthcare
• Wound Care Technology

Background:

• Chronic wounds pose a global health challenge, exacerbated by conditions like diabetes and Hansen's disease.
• Current visual inspection methods for wound tracking are hindered by accessibility issues in rural areas.
• There is a growing need for accessible and efficient wound monitoring solutions.

Purpose of the Study:

• To present a deep learning framework for chronic wound tracking using smartphone-captured RGB images.
• To integrate wound detection, segmentation, and quantitative analysis (area, perimeter) into a cohesive system.
• To introduce a novel dataset of chronic wounds from leprosy patients for research.

Main Methods:

• Development of a deep learning framework for processing RGB images of chronic wounds.
• Integration of established medical image processing algorithms for wound analysis.
• Utilizing smartphone cameras to capture wound images, eliminating the need for specialized equipment.

Main Results:

• The proposed framework demonstrates validity and accuracy in chronic wound tracking.
• Achieved up to 84.5% precision in experimental evaluations.
• Successful integration of wound detection, segmentation, area, and perimeter quantification.

Conclusions:

• The deep learning framework offers a viable, accessible solution for chronic wound monitoring, especially in underserved regions.
• The provided dataset will aid further research in chronic wound analysis and management.
• Smartphone-based imaging combined with AI shows promise for improving chronic wound care outcomes.