A Real-Time IoT and Machine Learning-based Smart Insole System for Early Detection and Prevention of Diabetic Foot Ulcers
R Rakshita, M. Lakshmanan, C. S. Karthik, Challa Venkata Sai, K Sai Dhanush, Matam Mahesh
Abstract
Diabetic Foot Ulcers (DFUs) are a severe and common complication of diabetes, affecting up to 50% of long-term patients and often leading to hospitalization or amputation. Existing monitoring methods lack affordability, scalability and real-time responsiveness. This study addresses the critical need for early detection and prevention of DFUs by developing a real-time, cost-effective smart insole system that leverages IoT and machine learning. The proposed system continuously monitors foot pressure using embedded Flexi-Force sensors and transmits data to a mobile application via a cloud-based platform. Advanced machine learning models, including K-Nearest Neighbors and Random Forest classifiers, analyze the data to predict early signs of ulceration. The system achieved a test accuracy of 99.997% in detecting abnormal pressure patterns, demonstrating its efficacy in identifying high-risk conditions before ulceration occurs. This work contributes a practical, lightweight, and reliable solution to diabetic foot care, improving patient outcomes through timely intervention.