Apparently random fluctuations in physiological time series may in fact be remarkably structured. Study of these fluctuations reveals that the current state of a neurally controlled system influences future states over an unexpectedly long period of time. Quantitative analysis of this far reaching influence indicates that such fluctuations are correlated over multiple time scales in a distinct mathematical pattern corresponding to long-range correlations with power-law decay. The general aim of the proposed research, therefore, is to investigate the unexplained, irregular dynamics of a wide class of complex physiological fluctuations under neural control-both autonomic and non- autonomic, by methods recently developed in statistical physics and non- linear mathematics. Furthermore, to study the alterations of long-range correlation properties under pathologic conditions as potential tools for neurophysiological monitoring. Two specific model systems will be analyzed for elucidating neurophysiological control: (i) fluctuations in human cardiac interbeat interval time series, and (ii) human gait interstride fluctuations. The former is under direct neuroautonomic control while the latter is non-autonomic. Both have the advantage that they can be readily measured non-invasively. I am applying for the FIRST Award to pursue independent investigations with the following specific aims: 1) Test the hypothesis that long-range correlations are generated by normal neurophysiological control mechanisms. 2) Test the hypothesis that alterations of long-range correlations indicate pathological perturbations and thus can be used for clinical monitoring. 3) Further develop and refine statistical methods that can detect and characterize long-range correlations in non-stationary physiological signals. 4) Develop theoretical models that account for the presence of long-range correlations associated with neurophysiological control mechanisms. These models should also be able to account for the breakdown of long-range correlations under certain pathological conditions.
These aims are directly related to the practical need for developing indices of neurophysiological disorders from non-invasive data as well as the long term goal of understanding important features common to neural control mechanisms.
