A Comparative Analysis of Deep Learning Models for Detection of Diabetic Foot Ulcer using Foot Thermography Images

Diabetic foot ulcers (DFUs) are a serious consequence for diabetes individuals, frequently resulting in amputation. Early identification is critical in averting such results. This study looks at the effectiveness of deep learning models in recognizing DFUs from foot thermography photos. We evaluate the performance of several deep learning architectures against more contemporary models. The models will be trained and assessed using a publicly accessible foot thermography picture collection that includes both healthy and DFU patients. Each model’s performance in DFU identification will be evaluated based on accuracy, precision, recall, and $\mathbf{f 1}$ score. We intend to find the deep learning model with the greatest accuracy,f1 score, and robustness in DFU classification, which would pave the way for a computer-aided diagnostic tool that uses thermography to detect DFU in diabetes patients early. This research shows potential for enhancing patient care and lowering amputation rates.