Detecting Diabetic Foot Ulcer with CNN and CapsNet: A Comparative Investigation of Model Performance

Abstract

Diabetic foot ulcer (DFU) poses a substantial health risk to patients and healthcare resources, necessitating accurate and efficient diagnosis. This study harnesses the power of convolutional neural networks (CNNs) and capsule networks (CapsNets) to redefine DFU classification, aiming for enhanced accuracy and efficiency. By integrating state-of-the-art deep learning architectures, we address the limitations of existing diagnostic methodologies. Through comprehensive data pre-processing, feature extraction, and classification, our system seeks to improve diagnostic precision and contribute to better patient outcomes. A comparative analysis between CNNs and CapsNets evaluates key performance metrics, guiding future advancements in medical image diagnosis. Furthermore, computational timing assessments ensure practicality in medical settings. This research underscores the potential of deep learning in revolutionizing medical image diagnosis and emphasizes the critical role of accurate DFU detection in diabetic care.