Human Movement Modeling
Mehdi Yousefi Azar Khanian; Seyed Mohammad Reza Hashemi Golpayegani; Mostafa Rostami
Volume 13, Issue 1 , April 2019, , Pages 55-68
Abstract
Recently, analysis of the human postural stability has gained increasing interest. This is mainly due to the necessity of understanding the self-organization mechanisms in this system activated in response to any motion pattern. The extraction of effective indicators from this system could help ...
Read More
Recently, analysis of the human postural stability has gained increasing interest. This is mainly due to the necessity of understanding the self-organization mechanisms in this system activated in response to any motion pattern. The extraction of effective indicators from this system could help clinicians to diagnose patients’ postural disorders and guide the rehabilitation processes. The center of pressure (CoP) signal, as a collective variable, contains information from the human equilibrium system. Through the CoP trajectory production, various control mechanisms are activated at different time intervals, which is equivalent with emerging different basin of attractors in the phase space. The dynamical coordination of this system patterns determines how system switches between these attractors. In this paper, first to quantify the local information of CoP, two indicators are defined; "local correlation dimension (LCD)" and "phase dynamic coordination (PDC)". Then, for a designed experiment, the local behavior pattern of CoP time series is calculated based on the suggested indicators. Next, by designing a model that can generate rich dynamics with multiple attractors, we attempt to follow data behavioral changes. The proposed model is map based. The model parameters are tuned by PCD to follow the pattern of sub-attractors changes with the system LCD. Tracking the behavioral patterns of the posture system is one of the prominent results of this research. The proposed model not only can follow the local behavior of system, but also follows the global dynamics. Accordingly, the similarity of the decreasing-increasing trend of the correlation dimension variations for the model output and data demonstrates the variations of system’s degrees of freedom in the test trials. The proposed model is the first behavioral model for the posture system, which can be used to quantify the variation of information in other biological systems based on the proposed methods.
