AthleTwin: Towards AI-Driven Digital Twins for Personalized Sports Injury Prevention and Rehabilitation

Abstract

Sports injuries impose substantial physiological, psychological, and economic burdens on athletes worldwide. Although artificial intelligence (AI) and machine learning (ML) have shown increasing potential for predicting injury risk, most existing approaches remain retrospective, population-based, and reliant on a single data modality, limiting their ability to capture the complex, individualized, and dynamic nature of sports injuries. Digital twin (DT) technology, which creates a continuously evolving virtual representation of a physical entity using real-time sensor data, offers a promising solution to address these limitations. This study presents AthleTwin, the first comprehensive conceptual framework for AI-driven digital twins specifically developed for sports injury prevention and rehabilitation. AthleTwin comprises five key players: (i) multi-modal data acquisition from wearable IoT sensors, computer vision, and medical imaging; (ii) personalized biomechanical and physiological modeling; (iii) AI-based injury risk prediction using hybrid physics-informed neural networks; (iv) an explainable AI (XAI) layer to support clinicians and coaches in decision-making; and (v) a closed-loop adaptive intervention engine for real-time load management and rehabilitation optimization. Further, it introduces a taxonomy of digital twin maturity levels in sports medicine, examine the major technical challenges spanning data, modelling, AI, and deployment, and outline a structured five-priority research agenda for future work. Thus, this work establishes a conceptual foundation and practical roadmap for the development of personalized, proactive, and clinically grounded AI systems for athlete health management.

Keywords