Nonlinear Signal Analysis
A major challenge in understanding biological systems is to connect the signaling on microscopic levels to the dynamics observed on the integrative, organ system level and functional outcome measures. Physiological recordings exhibit rich and complex dynamical patterns that contain high information content about functional state. Furthermore, loss of these complex dynamics is often correlated with a disease.
The following methods for quantification of complex system dynamics are available through collaboration with Dr. Goldbeger and Dr. C.K. Peng (Margret and H.A. Rey Institute for Nonlinear Dynamics at BIDMC/Harvard Medical School ( http://www.physionet.org).
Hilbert-Huang Transform
The Hilbert-Huang transform analysis,based on theories of chaotic systems, has been applied to
numerous physical and physiological systems to assess non-linear
relationship between two signals.
Multimodal Pressure Flow Method (MMPF)
The MMPF technique that is based on Hilbert-Huang transform, measures coupling between two
nonstationary signals. This method was implemented to quantify changes
in pressure-flow regulation (cerebral autoregulation) dynamics in patients with hypertension,stroke and diabetes using the Valsalva maneuver and spontaneous
pressure and flow oscillations.Cerebral autoregulation in healthy
subjects can be characterized by large phase delays between pressure and flow
oscillations. Decreased phase delays are associated with cerebrovascular
disease and strokes.
Synchronization Method
Phase synchronization analysis allows the exploration of nonlinear feedback interactions between different
physiological systems demonstrating complex oscillating behavior
(cardiovascular, gait and respiratory).
Multiscale Entropy (MSE)
Multiscale entropy method is useful for the complexity analysis of a variety of physiologic time series such as heart beat time series.
Modified Wigner distribution
Wigner distribution allows calculation of time-frequency distribution with high resolution. It was demonstrated that modified WD has useful properties for the
evaluation of nonstationary, short data series, i.e. in
electroencephalographic waves, heart rate, blood pressure, respiration,
cerebral blood flow, and other physiological signals.
Stabilogram-Diffusion analysis
Postural control is assessed from
the center of pressure (COP) displacements using stabilogram-diffusion analysis that provides dynamics measures of COP fluctuations. The COP signals behave as a
positively correlated random walk over short-time scales characterized
by a Hurst exponent.
Mathematical Modeling
Encounters among multiple variables and co-morbidities are common in
geriatric clinical research. In collaboration with Dr. M. Olufsen,
Dept. of Mathematics, North Carolina State Univ., we use
multicompartmental models to distinguish contributions of multiple
mechanisms enaged in short-term cardiovascular and cerebrovascular
control. The model is used to evaluate effects of aging and
hypertension on orthostatic adaptation to standing-up and to predict
variables that cannot be measured experimentally (i.. nerve firing,
baroreflex firing). Collaborative project with RBGHS group (http://www.uni-graz.at/biomath/Collaboration_group)(
Departments of Mathematics at the NC State Univ. (Dr. Olufsen, Dr.
Tran), Univ, Roskilde., Denmark (Dr. Ottesen), Univ.Gratz, Austria (Dr.
Kappel, Dr. Batzel), and the Safe Lab.
