Enhancing Diabetic Shoe Insole Customization with AI-Driven Approaches

The increasing prevalence of diabetes has led to a rise in complications like Diabetic Foot Ulcers (DFU) and neuropathic damage. Customized diabetic insoles are crucial for preventing these issues by redistributing pressure and providing support. Existing insole customization methods involve manual hole-cutting at the site of DFUs, lacking automation. This process is time-consuming and subjective, resulting in suboptimal outcomes. This research introduces “DiabiSole,” a web-based application that expedites wound area identification and measurement, detects elevated pressure areas on the foot, and delineates sites needing pressure offloading. Two Image Segmentation models, both belonging to the U-Net framework, were deployed to achieve wound identification and high-pressure region detection, achieving accuracies of 95% and 98%, respectively. The OpenCV library was utilized for quantifying relevant areas, and Pixel Distribution Analysis was employed to highlight areas requiring pressure offloading. “DiabiSole” is an innovative system in Sri Lanka that revolutionizes the insole customization for DFUs by replacing the laborious manual customization process. This system significantly improves patient outcomes by offering a personalized and precise approach while reducing future DFU occurrences. It addresses the critical need for efficient and accurate solutions in managing diabetic complications, ultimately enhancing the quality of life for those affected by diabetes.