Exploring Human Adaptation Using Ultrasonic Measurement Techniques and Optimized, Dynamic Human Neck Models

Abstract

Introduction: Classic inverse/dynamics methods fall short in providing a methodology which allows the human model to adapt to a change in environment or equipment. This paper introduces a dynamic human model with joint torques governed by coupled feedback controllers and optimized to a specific objective to allow for a structured approach in human adaptation simulation. Methods: Motion data from a data collection source (i.e., video, sensors, etc.) consists of 3D position histories of markers placed at various locations on the human subject. One such method involves new device employing an ultrasonic measurement method to capture body movements (orthoson). The device is completely self-contained and carried on the human body. It consists of a small central unit and sensors which are applied at specific locations on the skin. Changes in the distance between any pair of sensors are reported with an accuracy of a millimeter. By tracking the relative distances between markers through time, the data from this device can be used to drive a dynamic human model to simulate any human activity. [McGuan, 1994]. The ADAMSâ (Mechanical Dynamics Inc.) mechanical simulation system is used to process these data into a dynamic human model. Results: The simulation was performed using a dynamic human neck model. The process of simulation begins by positioning “motion agents” on the multisegmented human model at the same relative locations as in the experiment. The motion agents are driven with the experiment motion data. The agents are physically attached to the corresponding human segment using 6 degree-of-freedom (DOF) spring elements (bushings). The stiffness of this connection is normalized to the relative accuracy rating of the specific target marker in the experiment (available from the motion analysis equipment/software). This allows for the marker with a higher degree of accuracy to contribute more to the motion in the model changing the nature of the motion data to motion influencing rather that motion governing. A kinematic analysis is performed with this arrangement to retrieve the joint rotation histories for the human model emulating the motion. Conclusion: This combination of ultrasonic measurement techniques and forward dynamic simulation promises much utility in muscle control and coordination research, as well as practical applications of sports simulations, gait analysis, etc. References: McGuan, S., et al. (1994). An Approach for a Detailed Analytical Model of the Human Lower Extremity During a Drop Landing. ASB, Columbus, Ohio.