A Personalized Running Fatigue Monitoring System Based on IMU and PPG

Nowadays, exercise-related injuries, especially like knee damage, frequently occur due to improper intensity evalu- ation while doing physical activities. In order to solve this issue, this study presents a wearable system for real-time injury-risk prediction. Inertial measurement units (IMUs) and Photoplethys- mography (PPG) sensors are combined so that running data can be collected such as step frequency and HRV characteristics like Root Mean Square of Successive Differences and Standard Deviation of Normal-to-Normal Intervals (RMSSD/SDNN). These data are able to be used to analyze exercise intensity and to train a machine learning model(LSTM). Exercise intensity is divided into three levels: low, moderate, and high. When potentially risky patterns appear, such as unstable step frequency or reduced RMSSD/SDNN, the model can detect them and alert users (e.g., a red LED) to stop or adjust activity. This approach may help reduce injury risk.