This thesis explores the creation and validations of a simulated musculoskeletal model of the human hand with a focus on the aspects of... Show moreThis thesis explores the creation and validations of a simulated musculoskeletal model of the human hand with a focus on the aspects of pinching. Specifically, the thumb, index finger, and wrist were represented in OpenSim 3.3, using anatomical definitions for increased accuracy. Specifically, the inclusion of physiological axes of rotation at all joints, anatomically accurate passive joint torques, and appropriate moment arms for each muscle. The model was subsequently validated against experimental results found in literature. First, the digit tip force directions produced by each of the 15 muscles were compared to those obtained by loading the corresponding tendons in cadaveric specimens and measuring three-dimensional force generation at the tip of the thumb or index finger. Second, isometric force generation by activation of multiple muscles were compared. Finally, dynamic simulations were run using electromyographic (EMG) recordings as inputs. The capabilities of the model were then explored by using it to predict activation patterns from imposed movement and to simulate extension deficits in a hand affected by stroke. The model generated isometric force in the correct directions for most individual muscles, with the extensor pollicis brevis (EPB) showing the largest directional differences between cadaveric and simulated results. With combined muscle activation patterns the model simulated force profiles accurately, showing only a 5.3% mean squared error (MSE) from the actual force profile. In terms of force magnitudes between the model and simulated results, the model produced significantly lower force magnitudes, especially in the thumb. This validation was also found to be reasonably accurate to the expected motions. With the model anatomically validated, two different simulations were run using the model. First, known kinematics were applied to the model and the muscle activations were simulated; the resultant joint angles were found to match the expected within 10% MSE. Second, a stroke affected hand was simulated, with activation deficits added to each of the muscles individually. It was found through this that, in the model, the intrinsic muscles played a larger role in force production and dynamic motion than the extrinsic muscles. In all, these validations and simulations produce a promising groundwork for the use of this model for further simulation. M.S. in Biomedical Engineering, May 2016 Show less
