Wearable Smart Insole for Early Diabetic Foot Ulcer Prediction Using Multi-Modal Sensing

Diabetic foot ulcers (DFUs), a chronic complication of diabetes, accounts for severe infections, more than 85% of diabetes-related amputations, and increased mortality rate, if not detected and managed at an early stage. Continuous foot health monitoring of diabetes patients is critical to prevent severe outcomes. In this paper, a smart insole system is designed for continuous monitoring of vital foot health parameters via integrated multi-sensor technology. Six pressure sensors FSR 402, two temperature and humidity sensors SHT31, and one pulse oximeter sensor MAX30102, are mounted on the insole in accordance with the clinically relevant, high-risk areas. Data from these sensors is wirelessly communicated to the monitoring graphical user interface (GUI), offering real-time visualization and alerts. Performance of this insole was then evaluated on a set of individuals with different BMI ranges, in different postures and positions, and a versatile database was collected. A Long Short-Term Memory (LSTM) model is trained on the acquired dataset, which then predicts distinct pressure changes on six DFU prone areas, localized temperature variations, and peripheral blood oxygen levels of individuals and alerts them about the possible likelihood of an ulcer. This validates the insole's utility as an early proactive monitoring device for DFU detection and prevention.