A VISUAL DATASET OF DIABETIC FOOT ULCERS FOR AUTOMATED DETECTION USING COMPUTER VISION

DATASET VISUAL ULKUS KAKI DIABETIK UNTUK DETEKSI OTOMATIS MENGGUNAKAN COMPUTER VISION

Abstract

Background: Diabetic foot ulcers (DFUs) are a common complication of diabetes mellitus that require continuous monitoring, particularly in homecare settings. Computer vision offers promising tools for automating wound assessments. This study aimed to develop a visual dataset of diabetic foot ulcers to support future computer vision applications for automatic wound detection. Methods: A total of 30 patients with diabetes (13 men and 17 women) receiving homecare services at Kartika Husada Clinic, Malang Regency, and its affiliated group were included. Wound images were collected over four months using standardized photography. Each image documented the location, size, tissue condition, and signs of infection. Results: Most wounds were located on the plantar or heel areas. The visual characteristics included necrotic tissue, granulation, purulent discharge, and signs of inflammation. The dataset reflects variations in wound appearance and severity, which are essential for developing robust machine learning models. Conclusion: The resulting dataset provides a foundation for developing automated wound-detection tools using computer vision. This also supports the implementation of intelligent monitoring systems.