Advanced Classification of Diabetic Foot Ulcer Ischemia and Infection Utilizing Efficient Net Deep Learning Models

Diabetic foot ulcers are long-standing deterioration of DM and cause severe ischemia and infection that may lead to an amputation if left unrecognized. This work introduces CNN Efficient Net Deep learning as an improved classifier for the identification of DFU ischemia and infection. To train the model the study uses an extensive dataset of DFU images which have been annotated for enhancement with the help of Efficient Net that provides a compound scaling in terms of performance. Preprocessing is applied on images for improvement in quality of images and then the trained model is tested and validated for generalization to different clinical scenarios. This is accomplished by maintaining the appropriate depth, width, and resolution for efficient classification within the model, while also overcoming information bottlenecks in the feedforward neural network, all of which is made possible by Efficient Net architecture. The study shows promising performance of the proposed models and emphasizes the impact on classification metrics compared organized fashion rather than time-based traditional approaches, that is inherent in clinical decision-making process. Moreover, it also reveals that deep learning can potentially help the healthcare provider identify and address the DFCs accurately. Finally, our results corroborate the commitment to using artificial intelligence as a practice changer in the delivery of diabetic foot care to decrease the chances of amputations among diabetic patients.