Related Experiment Videos
A statistical mechanical analysis of postural sway using non-Gaussian FARIMA stochastic models
1angelo@helios.sssup.it
IEEE Transactions on Bio-Medical Engineering
|September 29, 2000
Summary
This study models postural sway using fractional ARIMA models, viewing center-of-pressure motion as an anti-persistent random walk. Vision significantly impacts model parameters like the Hurst exponent in healthy adults.
Area of Science:
- Biomechanics
- Dynamical Systems
- Statistical Modeling
Background:
- Postural sway is a complex physiological process crucial for balance.
- Existing models often simplify the underlying dynamics of center-of-pressure (COP) motion.
- Understanding COP dynamics is vital for assessing stability and neurological function.
Purpose of the Study:
- To model postural sway using a fractional autoregressive integrated moving average (FARIMA) framework.
- To characterize COP motion as a self-similar, anti-persistent random-walk process.
- To investigate the influence of visual input on postural sway model parameters.
Main Methods:
- Utilized a FARIMA family of models to represent COP motion.
- Viewed COP motion as a fractionally summated, anti-persistent random-walk process.
- Employed a graphical Hurst exponent estimator and higher-order cumulant-based AR model fitting.
Main Results:
- The model successfully captures the correlation structure of COP motion, influenced by a low-pass filter.
- Identified key parameters: stochastic driving strength, filter characteristics (DC gain, damping, natural frequency), and Hurst exponent.
- Demonstrated that the presence or absence of vision alters the estimated model parameter values in healthy young adults.
Conclusions:
- FARIMA models provide a robust framework for analyzing postural sway and COP dynamics.
- The Hurst exponent effectively quantifies the anti-persistence magnitude in random-walk processes of sway.
- Visual feedback plays a significant role in modulating the parameters governing postural control.