Severity of Diabetic Foot Ulcer Patients Prediction using a Hybrid and Machine Learning Techniques

The Diabetic Foot Ulcers (DFUs) are one of the most disabling complications of diabetes mellitus conditions which most of the times result in the course of infection, amputation, and fatality in case of failure management. Accurate forecasting of the severity of DFU in a patient at their initial phases is a key milestone in ensuring a timely response on part of the clinical fraternity and improved patient only outcomes. DF competed with DL and ML methods have become an effective measure of DFU risk assessment, but some limitation remain with the use of the two methods in their interpretability, generalizability, and accuracy. Survey in this review is accompanied by the systematic review of the current trends in DFU severity prediction, especially with the focus on hybrid and novel ML algorithms. We explore the data sets, data pre processing methods, data features selection methods, machine learning techniques and measurement boxes. Further, new advances are also critically evaluated in terms of hybrid models that integrate classical ML with the deep learning, fuzzy systems or optimization algorithms. The present limitations, clinical applicability and future research direction form the conclusion of the review to make attempts to bridge the opportunity between intelligent systems and personalized DFU management.