Robust Real-Time Musculoskeletal Modeling Driven by Electromyograms

Abstract

Abstract-Objective: Current clinical biomechanics involves lengthy data acquisition and time-consuming offline analyses. Current biomechanical models cannot be used for real-time control in man-machine interfaces. We developed a method that enables online analysis of neuromusculoskeletal function in vivo in the intact human. Methods: We used electromyography (EMG)-driven musculoskeletal modeling to simulate all transformations from muscle excitation onset (EMGs) to mechanical moment production around multiple lower-limb degrees of freedom (DOFs). We developed a calibration algorithm that enables adjusting musculoskeletal model parameters specifically to an individual's anthropometry and force-generating capacity. We incorporated the modeling paradigm into a computationally efficient, generic framework that can be interfaced in real-time with any movement data collection system. Results: The framework demonstrated the ability of computing forces in 13 lower-limb muscle-tendon units and resulting moments about three joint DOFs simultaneously in real-time. Remarkably, it was capable of extrapolating beyond calibration conditions, i.e. predicting accurate joint moments during six unseen motor tasks and about one unseen DOF. Conclusion: The proposed framework can dramatically reduce evaluation latency in current clinical biomechanics and open up new avenues for establishing prompt and personalized treatments, as well as for establishing natural interfaces between patients and rehabilitation systems. Significance: The integration of EMG with numerical modeling will enable simulating realistic neuromuscular strategies in conditions including muscular/orthopedic deficit, which could not be robustly simulated via pure modeling formulations. This will enable translation to clinical settings and development of healthcare technologies including real-time bio-feedback of internal mechanical forces and direct patient interfacing with wearable assistive devices.

Approximating complex musculoskeletal biomechanics using multidimensional autogenerating polynomials

“…The complex geometric interactions-sliding and wrapping-between muscles and other mechanical body structures pose a considerable computational challenge for real-time applications (Blana et al, 2017). The engineering trade-off between complexity, performance, and accuracy pushed the development of simplified biomechanical limb models that assumed constant moment arm and posture relationships (Crouch and Huang, 2016) or reduced the span of musculotendon anatomy to ease computational demand (Durandau et al, 2018). Here, we report a method of capturing the kinematic MS transformations within the biomechanical model of the forearm and hand that does not require these simplifications.…”

“…Interest in MS approximations has been steadily increasing with the development of computational tools for human motion analysis, e.g., OpenSim (Delp et al, 2007). Accuracy of these approximations has been demonstrated with B-spline models (Durandau et al, 2018;Sartori et al, 2012) and computational efficiency has been achieved with polynomial models (Chadwick et al, 2009;Menegaldo et al, 2004). The optimal polynomials derived in this manuscript have the benefits of both accuracy and computational efficiency.…”

“…Another approach developed by Sartori and colleagues (Sartori et al, 2012) emphasizes the quality of approximation using cubic splines. Albeit being computationally expensive, the ability of this approach to operate at realtime has been shown in a 3-DOF per muscle model (Durandau et al, 2018). The drawback of cubic splines, however, are their limited scalability: the number of spline coefficients increases exponentially with the number of DOFs that the muscle crosses.…”

High muscle co‐contraction does not result in high joint forces during gait in anterior cruciate ligament deficient knees

“…While muscle co‐contraction indices (or more appropriately, co‐activation indices) can be computed using different methodologies, using these indices alone to predict joint loading may lead to erroneous conclusions. When estimation of joint loading is not possible, co‐contraction indices that are based on muscle joint moments (instead of muscle activation) may provide a better alternative . From a clinical perspective, decreased knee joint loading (and not excessive loading) at early post‐injury time points has been associated with knee osteoarthritis (OA), at least in a population that opts for reconstructive surgery .…”

“…When estimation of joint loading is not possible, cocontraction indices that are based on muscle joint moments (instead of muscle activation) may provide a better alternative. 19,34,35 From a clinical perspective, decreased knee joint loading (and not excessive loading) at early post-injury time points has been associated with knee osteoarthritis (OA), at least in a population that opts for reconstructive surgery. 36 For rehabilitation, evaluating knee joint loading can aid in determining the success of non-operative management in delaying the OA progression, and inform guidelines about returning to sport.…”