2024/07/20 by Karimzadeh, Mohammadsadegh, Naderi Nasab, Mehdi, Taheri, Morteza +1
Engineering · Health Professions · Medicine · #Artificial intelligence #Muscle activation and electromyography studies #Occupational Health and Performance #Sports Performance and Training #biomechanics #machine learning #military training #motion analysis #physical performance optimization
paper · doi:10.71626/anmd.2024.1221007
openalex publication_date 2024/07/20 · openalex created_date 2025/11/05 · openalex updated_date 2026/07/22
With advances in machine-learning methods, fusing biomechanical data with artificial intelligence has become an efficient approach for motion analysis and training optimization. This study set out to develop and evaluate an intelligent system for biomechanical analysis and optimization of physical training among personnel of the Islamic Republic of Iran Army. Motion data collected during a battery of standard military exercises were recorded using inertial measurement units (IMUs) alongside synchronized video. After preprocessing, biomechanical features—including joint angles, angular velocity and acceleration, and ground reaction forces (GRF)—were extracted. To identify movement patterns and assess performance indices, AI models comprising deep neural networks (DNN/CNN–LSTM) and support vector machines (SVM) were employed. Results showed that the system achieved accuracy >92% in distinguishing optimal movements from inefficient patterns associated with increased joint loading and muscular fatigue. Incorporating the system’s outputs into personalized training prescriptions yielded, in pre–post evaluations, an 18% reduction in the estimated risk of musculoskeletal injury and a 15% improvement in physical performance indices. Overall, the findings indicate that integrating AI and biomechanics offers an effective pathway to intelligent military training, enhanced combat readiness, and reduced training-related injuries across the armed forces