Force Closure Mechanism Modeling for Musculoskeletal Multibody Simulation

• Published in: IEEE transactions on biomedical engineering Vol. 65; no. 11; pp. 2471 - 2482
• Main Authors: Geier, Andreas, Aschemann, Harald, D'Lima, Darryl, Woernle, Christoph, Bader, Rainer
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.11.2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Biological system modeling; Biomechanics; Computational modeling; Computer simulation; Contact force; Decision making; Dynamic models; Dynamics; Equations of motion; Fasteners; Force; instantaneous screw axis; joint modeling; Kinematics; Knee; Mathematical model; mechanical systems; Multibody systems; Muscles; Musculoskeletal diseases; musculoskeletal multibody system; Noise prediction; parameter identification; Parameterization; Pendulums; Predictive control; rehabilitation robotics
• ISSN: 0018-9294; 1558-2531; 1558-2531
• DOI: 10.1109/TBME.2018.2800293

Objective: Neuro-musculoskeletal multibody simulation (NMBS) seeks to optimize decision-making for patients with neuro-musculoskeletal disorders. In clinical practice, however, the inter-subject variability and the inaccessibility for experimental testing impede the reliable model identification. These limitations motivate the novel modeling approach termed as force closure mechanism modeling (FCM 2 ). Methods: FCM 2 expresses the dynamics between mutually articulating joint partners with respect to instantaneous screw axes (ISA) automatically reconstructed from their relative velocity state. Thereby, FCM 2 reduces arbitrary open-chain multibody topologies to force closure n-link pendulums. Within a computational validation study on the human knee joint with implemented contact surfaces, we examine FCM 2 as an underlying inverse dynamic model for computed muscle control. We evaluate predicted tibiofemoral joint quantities, i.e., kinematics and contact forces along with muscle moment arms, during muscle-induced knee motion against the classic hinge joint model and experimental studies. Results: Our NMBS study provided the proof-of-principle of the novel modeling approach. FCM 2 freed us from assuming a certain joint formulation while correctly predicting the joint dynamics in agreement with the established methods. Although experimental results were closely predicted, owing to noise in the ISA estimation, muscle moment arms were overestimated (RISA = 0.84 <; RHINGE = 0.97, RMSEISA = 13.18 mm > RMSEHINGE = 6.54 mm), identifying the robust ISA estimation as key to FCM 2 . Conclusion: FCM 2 automatically derives the equations of motion in closed form. Moreover, it captures subject-specific joint function and, thereby, minimizes modeling and parameterization efforts. Significance: Model derivation becomes driven by quantitative data available in clinical settings so that FCM 2 yields a promising framework toward subject-specific NMBS.