An open-source model and solution method to predict co-contraction in the index finger

Abstract

Determining tendon tension in the finger is essential to understanding forces that may be detrimental to hand function. Direct measurement is not feasible, making biomechanical modelling the best way to estimate these forces. In this study, the intrinsic muscles and extensor mechanism were added to an existing model of the index finger, and as such, it has been named the Intrinsic model. The Intrinsic model of the index finger has 4 degrees of freedom and 7 muscles (with 14 components). Muscle properties and paths for all extrinsic and intrinsic muscles were derived from the literature. Two models were evaluated, the Intrinsic model and the model it was adapted from (identified in this thesis as the Extrinsic-only model). To complement the model, multiple static optimization solution methods were also developed that allowed for EMG-constrained solutions and applied objective functions to promote co-contraction. To test the models and solution methods, 10 participants performed 9 static pressing tasks at 3 force levels, and 5 free motion dynamic tasks at 2 speeds. Kinematics, contact forces, and EMG (from the extrinsic muscles and first dorsal interosseous) were collected. For all solution methods, muscle activity predicted using the Intrinsic model was compared to activity from the model currently available through open-source software (OpenSim). Just by using the Intrinsic model, co-contraction increased by 16% during static palmar pressing tasks. The EMG-constrained solution methods gave a smaller difference between predicted and experimental activity compared to the optimization-only approach (p < 0.03). The model and solution methods developed in this thesis improve co-contraction and tendon tension estimates in the finger. As such, this work contributes to our understanding of the control of the hand and the forces that may be detrimental to hand function.