A Deterministic Model of Human Motion Based on Algebraic Techniques and a Sensor Network to Simulate Shoulder Kinematics.

Abstract

Limiting the quantitative characterization of ambulatory mobility to only the two-dimensional sagittal plane through the investigation of key kinematic parameters, may still inform scientists and bioengineers of critical elements of joint locomotion. This paper presents the initial validation of a deterministic biomechanical gait model that was derived from an inverse kinematic analysis of three-dimensional upper extremity movement. Algebraic methods were applied to generate shoulder flexion and extension angles during a single gait cycle during normal walking. The direct kinematic measurements from a motion capture system were analyzed and compared to the predicted measurements from the algebraic model for eight healthy, human subjects. The predicted results over all subjects varied from the actual joint angle measurements with a nominal amount of mean error (23%), while correlations were quite strong (mean R² = 0.97). These findings indicate the potential value of deterministic modeling with algebraic techniques as an alternative to existing methods.