Deep Learning Approaches for Diabetic Foot Ulcer Diagnosis and Classification Across Various Imaging Modalities

A dangerous and common consequence among diabetes patients, diabetic foot ulcers (DFU) can result in infection, amputation, and higher healthcare costs if they are not detected and treated promptly. As imaging technology become more widely available and artificial intelligence (AI) advances, research into the automated detection and classification of DFUs has accelerated. Using a variety of imaging modalities, such as RGB imaging, thermal and infrared imaging, hyperspectral imaging, ultrasound, X-ray, and magnetic resonance imaging (MRI), this literature review offers a thorough summary of current approaches for DFU analysis. In order to tackle problems including ulcer segmentation, tissue categorization, severity grading, and bone infection (osteomyelitis) identification, the study investigates traditional machine learning, deep learning, and hybrid computational approaches. Particular focus is placed on robustness to clinical variation and distribution shifts, interpretability of the model, and multimodal fusion approaches. Persistent issues such false correlations, inter-patient variability, limited labeled datasets, and model generalizability across clinical settings are also covered in the survey. This work attempts to facilitate the development of more efficient, scalable, and clinically useful DFU diagnostic tools by summarizing the most recent research trends, technological developments, and observed limitations.