PROQUEST: 1904509774
The long term goal of this thesis is to create quantitative, clinically significant measures that allow for early detection of Parkinson’s disease (PD) postural instability (PI), the progression of PI due to PD progression, and ultimately, fall risk in PD patients. Current clinical assessments in PD are not sufficiently sensitive to predict fall risk. Although biomechanical postural sway measures have provided quantitative characterization towards the progression of PI associated with PD progression, these methods are still not sufficiently sensitive to allow for early detection of PD and fall risk. Thus, a need arises for new quantitative methods to be established which can further describe PI progression in PD. This thesis had two overall goals:
Specific Aim 1 determined through iterative testing of input parameter combinations that both AFA and DFA are highly sensitive to input parameters when considering fractional Brownian motion (fBm) signals. Input parameter ranges for fBm-like signals in appropriately-large biological data should be examined at maximum window sizes (nmax) values between N/6 and N/10; minimum window sizes (nmin) values around 4 to 6 samples; and for fitted polynomial order (M) for AFA to remain first order.
Specific Aim 2 showed that fractal analysis methods may be sensitive towards detecting the development and progression of PI in PD. AFA and DFA were tested on postural sway data collected in a previous study that used mild PD patients (Hoehn and Yahr stage (H&Y) 2, without postural deficits), moderate PD patients (H&Y 3, with postural deficits), and agematched healthy controls (HC). AFA produced the most clinically significant measure, Hfast, which detected changes in COPv dynamics across smaller time scales than other parameters. These results suggest that components of fractal analysis on COPv time series could be used in concert with traditional quantitative and clinical measures to further enhance the sensitivity of clinical analysis, the understanding of PD PI dynamics and progression, and development of predictive computational simulations of motor and postural control in PD.