End-User Skin Analysis (Moles) through Image Acquisition and Processing System.

Area of Science:

• Biomedical Engineering
• Dermatology
• Medical Imaging

Background:

• Early detection of skin cancer, such as melanoma, is crucial for effective treatment.
• Traditional mole analysis relies on visual inspection by medical professionals.
• A need exists for accessible, user-friendly tools for regular skin lesion monitoring at home.

Purpose of the Study:

• To develop an end-user device for capturing, segmenting, and diagnosing skin moles using the ABCD rule.
• To create a portable hardware solution with controlled lighting for consistent image acquisition.
• To develop PC software for image storage, tracking, and automated ABCD rule analysis.

Main Methods:

• A compact, handheld device with integrated LEDs was designed for easy manipulation.
• PC software was developed for local image database storage and analysis.
• Image processing algorithms were created to evaluate mole asymmetry, border, color, and diameter (ABCD rule).
• The system was validated using the international PH2 database and a custom database of 26 patient images with specialist diagnoses.

Main Results:

• The developed device achieved high performance in mole image analysis.
• Key performance metrics included Accuracy: 0.92, Precision: 1.0, Recall: 0.92, and F1-score: 0.96.
• The system demonstrated effective segmentation and criteria evaluation for suspicious moles.

Conclusions:

• The developed end-user device offers a promising tool for regular, at-home monitoring of skin moles.
• The system supports early identification of potentially cancerous lesions by applying the ABCD rule.
• This technology can aid individuals in tracking mole changes and facilitate timely consultation with medical specialists.