Accurate estimation of muscle forces is essential for understanding biomechanics, injury risk, and rehabilitation strategies. This study develops a correlation-based regression model in predicting muscle forces from surface electromyography (sEMG) data and compares it with estimations obtained from OpenSim. Data obtained from individuals clustered into seven weight categories were analyzed, with a key focus on lower-limb muscles: rectus femoris (RF), vastus lateralis (VL), vastus medialis (VM), biceps femoris (BF), and semitendinosus (ST). A total of 45 participants were recruited for this study. The participants performed maximal voluntary isometric contractions (MVIC) and submaximal voluntary isometric contractions (Sub-MVIC) from which, the sEMG data was obtained and the corresponding muscle force data was deduced. Regression equations were obtained linking the muscle forces to those of sEMG data, among various weight categories for all the five muscles. The results demonstrated a strong agreement between regression-based predictions and OpenSim simulations, reinforcing the validity of the present method. VL and VM exhibited the highest R 2 values ranging from 0.80 to 0.97, with near-exact force trend replicating exponential curves. Whereas RF, BF, and ST, being bi-articular muscles, have significant linear fits with R 2 ranging from 0.81 to 0.96. Interestingly, force slopes decreased progressively across higher weight categories, indicating lower muscle activations in individuals with higher body weight—a trend that may be linked to muscle fiber lengths. The present approach provides an efficient alternative to musculoskeletal modeling while maintaining good accuracy, making it suitable for performance analysis, rehabilitation, and clinical diagnostics. The strong correlation between regression-based and OpenSim-derived force predictions highlights the robustness of this method, making it a valuable tool in biomechanics research and clinical applications.
Teja et al. (2026) studied this question.