Mobile Application for Diabetic Foot Ulcer Detection

This article presents the development of a mobile application aimed at healthcare professionals, which uses Convolutional Neural Networks to identify injuries in patients with Diabetic Foot Ulcers. In addition to detailing the development methodology and presenting the prototype and the CNN model, the study also includes an evaluation by healthcare professionals. Three different YOLO versions were examined, and YOLOv8 was selected as the most suitable. The results reveal that the model achieved a Mean Average Precision of 89.90%, demonstrating its effectiveness in the detection. Furthermore, healthcare professionals praised the usability and intuitiveness of the system, suggesting small layout changes and additional features to improve its usefulness.