Nonlinear time series analysis of normal and pathological human walking

Area of Science:

• Biomechanics
• Neuroscience
• Human Movement Science

Background:

• Understanding neuromuscular control of locomotion is crucial, especially for fall risk assessment.
• Traditional methods for quantifying walking stability lack precision in assessing the neuromuscular system's response to perturbations.
• More refined definitions and methods are needed to accurately measure dynamic stability during human locomotion.

Purpose of the Study:

• To quantify local dynamic stability and scaling structure of human walking kinematics using novel methods.
• To compare dynamic stability during overground versus treadmill walking in healthy young adults.
• To investigate dynamic stability differences between diabetic neuropathic patients and healthy controls during overground walking.

Main Methods:

• Estimation of average maximum finite-time Lyapunov exponents to quantify local dynamic stability.
• Calculation of local scaling exponents to assess the local scaling structure of gait time series.
• Development of a modified surrogate data method to analyze the stochastic nature of gait fluctuations.

Main Results:

• Motorized treadmill walking led to statistically significant, albeit small, artificial stabilization of locomotor kinematics compared to overground walking.
• Diabetic neuropathic patients exhibited more locally stable gait patterns when reducing walking speed, resolving a paradox between variability and stability.
• Peripheral sensory loss in neuropathic patients correlated with distinct local scaling structures in their walking kinematics.

Conclusions:

• Novel methods provide a clearer understanding of dynamic stability and scaling in human locomotion.
• Findings resolve previous inconsistencies in the literature regarding gait variability and dynamic stability.
• The study highlights the role of sensory feedback in locomotor control and offers insights into the gait of individuals with diabetic neuropathy.