Development and Validation of an AI-Based Wound Alert System With a Home-Based Management Model for a Diabetic Foot Recurrence Cohort

Diabetic foot ulcer (DFU) is a major adverse outcome of diabetes, which itself is one of the most significant chronic diseases. The recurrence of DFU involves multiple risk factors, including altered foot loading patterns, patient compliance, family care capacity, blood glucose monitoring, degree of ischemia, and systemic disease control. Early identification of recurrence signs and timely follow-up interventions are crucial for improving prognosis, reducing disability rates, and lowering healthcare costs. However, traditional follow-up systems lack individualized strategies-such as risk stratification, inflexible follow-up intervals, and insufficient compliance management-often resulting in suboptimal outcomes. High-risk patients prone to recurrence may not be followed up frequently enough for early detection, while low-risk patients may undergo unnecessary visits, increasing burdens on both patients and healthcare providers. This inefficiency contributes significantly to the persistently high rates of disability and mortality among recurrent DFU patients.

Establishing an individualized follow-up strategy for DFU, supported by advanced technology to address core bottlenecks such as delayed recurrence warnings and inadequate home-based management, represents an effective technical pathway to tackle these issues.

Our center proposes to develop a dedicated DFU cohort with comprehensive active follow-up and a multimodal database encompassing well-defined indicators. We aim to explore a high-risk foot grading system for preventing DFU recurrence and design targeted follow-up protocols. By leveraging AI technology, we intend to build a wound warning system capable of identifying DFU recurrence. Furthermore, we seek to establish a telemedicine and AI-assisted, patient-centered home-based self-management framework for early warning and prevention of DFU recurrence.