Fractal Analysis of Center of Pressure Velocity Time Series in Parkinson's Disease

Abstract

Abstract The purpose of this study was to test the sensitivity of system parameters of the Center of Pressure velocity (COPv) time series using Detrended Fluctuation Analysis to pre-clinical postural instability (PI) in PD, the progression of PI due to PD progression, and ultimately fall risk. The long term goal is to create quantitative clinically significant measures of pre-clinical PD PI, the progression of PI due to PD progression, and fall risk. Postural sway data collected in a previous study, including participants with mild PD (PD-Mi), moderate PD (PD-Mo) and age-range-matched healthy controls (HC), were analyzed in this study. Ground reaction forces and moments were collected from subjects standing on force plates in quiet postural sway in eyes open (EO) and eyes closed (EC) conditions. COPv was calculated and analyzed as a non-stationary time series. We investigated the temporal parameter of Absolute Average Maximal Velocity (AAMV), the system order parameter of Approximate Entropy (ApEn), and fractal parameters from the DFA which were the short (α1) and long (α2) term scaling behavior of the time series and the time scale at which the behavior changes – the crossover index (CrI). AAMV showed significant group differences between HC and PD-Mo and significant condition differences. In the fractal analysis, α1 showed significant group differences between HC and PD-Mo and α2 showed significant differences between conditions. Due to the pilot nature of the study, power analysis was conducted on all non-significant measures in order to investigate required subject numbers for significance. Feasible subject numbers were found for many of the measures. These results suggest that the temporal and fractal analysis of the COPv time series are sensitive measures of the differences between PD and HC and can be used in concert with traditional measures to further benefit clinical analysis, understanding of disease pathology, and development of computer simulation models of postural control in PD.