Automated Detection of Diabetic Foot Ulcers using Multi-Level Thermal Image Data

A rise in diabetes globally has given rise to the condition of diabetic foot ulcers, which, when not treated early, can be a gateway to severe complications such as amputations. Early diagnosis and intervention are essential in enhancing patient outcomes; however, conventional diagnosis methods are mainly subjective and time-consuming. This work introduces an automated diabetic foot ulcer recognition system based on multi-level thermographic imaging data. The system proposed in this work takes thermal images at different angles and conditions that the CNNs use to extract various features and classify theulcers with a high accuracy. The model, trained on a diversified dataset of healthy and ulcerated foot thermal images, shows better performance metrics such as accuracy, sensitivity, and specificity compared with the conventional methods. This new system is efficient in time and minimizes errors caused by humans. This challenges the subjective assessment in clinical diagnosis. With this non-invasive thermographic imaging and machine learning algorithms, technology has the potential to revolutionize diabetic foot ulcer management, allowing timely interventions and improving patient outcomes. Future work will focus on refining the model and integrating the system into real-world clinical environments to further test its effectiveness.