Segmentation and classification of skin burn images with artificial intelligence: Development of a mobile application

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Mobile Health

Background:

• Accurate burn assessment is critical for effective treatment and patient outcomes.
• Accessible tools for immediate first aid guidance are needed, especially in remote areas.
• Artificial intelligence offers potential for automated analysis of medical images.

Purpose of the Study:

• To develop and evaluate an AI-driven mobile application for skin burn image analysis.
• To enable segmentation, classification, and object detection of burn wounds.
• To provide users with burn degree determination and first aid instructions.

Main Methods:

• Utilized the YOLO V7 architecture for image segmentation, classification, and object detection.
• Created a comprehensive burn image dataset, incorporating open-access sources and hospital data.
• Developed the Burn Wound Detection Android mobile application for real-world testing and accuracy assessment.

Main Results:

• The YOLO V7 model achieved a 75.12% success rate on the test dataset.
• The developed Burn Wound Detection application demonstrated accurate detection of burn images.
• The study successfully integrated AI for burn wound analysis and intervention guidance.

Conclusions:

• The AI-powered mobile application facilitates rapid burn degree assessment and first aid recommendations.
• This technology represents a significant advancement for public health, empowering individuals with critical burn care information.
• The developed system enhances accessibility to timely and accurate burn management guidance.