Prediction of diabetic foot ulceration using spatial and temporal dynamic plantar pressure

Diabetes Mellitus is a serious global health concern affecting about 415 million or 8.8% of adults worldwide. Among other complications of the disease, diabetic foot ulceration (DFU) is one of the most serious, possibly leading to amputation. This study presents the design, implementation and testing of a method for predicting DFU based on dynamic pressure distribution. We recorded the dynamic plantar pressure measurements during normal gait for 56 diabetic patients with and without diabetic peripheral neuropathy and 28 control non-diabetic subjects. Defining newly extracted features, employing machine learning techniques and applying Support Vector Machine classifier, achieved a classification accuracy and precision of more than 94.6% and 95.2% respectively. These promising results show the potential of the proposed method for predicting DFU allowing for early treatment and the possibility of providing diabetic patients with proper off-loading footwear to redistribute plantar pressure.