Search

A rich and complex temporal structure of variability in postural sway characterizes healthy and adaptable postural control. However, neurodegenerative disorders such as Parkinson's disease, which often manifest as tremors, rigidity, and bradykinesia, disrupt this healthy variability. This study examined postural sway in young and older adults, including individuals with Parkinson's disease, under different upright standing conditions to investigate the potential connection between the temporal structure of variability in postural sway and Parkinsonism. A novel and innovative method called oriented fractal scaling component analysis was employed. This method involves decomposing the two-dimensional center of pressure (CoP) planar trajectories to pinpoint the directions associated with minimal and maximal temporal correlations in postural sway. As a result, it facilitates a comprehensive assessment of the directional characteristics within the temporal structure of sway variability. The results demonstrated that healthy young adults control posture along two orthogonal directions closely aligned with the traditional anatomical anteroposterior (AP) and mediolateral (ML) axes. In contrast, older adults and individuals with Parkinson's disease controlled posture along suborthogonal directions that significantly deviate from the AP and ML axes. These findings suggest that the altered temporal structure of sway variability is evident in individuals with Parkinson's disease and underlies postural deficits, surpassing what can be explained solely by the natural aging process. [ABSTRACT FROM AUTHOR]

We propose a novel class of mixed fluctuations with different orientations and fractal scaling features as a model for anisotropic two-dimensional (2D) trajectories hypothesized to appear in complex systems. Furthermore, we develop the oriented fractal scaling component analysis (OFSCA) to decompose such mixed fluctuations into the original orientation components. In the OFSCA, the original orientations are detected based on the principle that the original angles are orthogonal to the angles with the minimum and maximum scaling exponents of the mixed fluctuations. In our approach, the angle-dependent scaling properties are estimated using the Savitzky–Golay-filter-based detrended moving-average analysis (DMA), which has a higher detrending order than the conventional moving-average-filter-based DMA. To illustrate the OFSCA, we demonstrate that the numerically generated time-series of mixed fractional Gaussian noise (fGn) processes with non-orthogonal orientations and different scaling exponents is successfully decomposed into the original fGn components. We demonstrate the existence of oriented components in the 2D trajectories by applying OFSCA to real-world time-series, such as human postural fluctuations during standing and seismic ground acceleration during the great 2011 Tohoku-oki earthquake. [ABSTRACT FROM AUTHOR]