Statistical quantifiers of memory for an analysis of human brain and neuro-system diseases

On the basis of a memory function formalism for correlation functions of time series we investigate statistical memory effects by the use of appropriate spectral and relaxation parameters of measured stochastic data for neuro-system diseases. In particular, we study the dynamics of the walk of a patient who suffers from Parkinson’s disease (PD), Huntington’s disease (HD), amyotrophic lateral sclerosis (ALS), and compare against the data of healthy people (CO — control group). We employ an analytical method which is able to characterize the stochastic properties of stride-to-stride variations of gait cycle timing. Our results allow us to estimate quantitatively a few human locomotion function abnormalities occurring in the human brain and in the central nervous system (CNS). Particularly, the patient’s gait dynamics are characterized by an increased memory behavior together with sizable fluctuations as compared with the locomotion dynamics of healthy patients. Moreover, we complement our findings with peculiar features as detected in phase-space portraits and spectral characteristics for the different data sets (PD, HD, ALS and healthy people). The evaluation of statistical quantifiers of the memory function is shown to provide a useful toolkit which can be put to work to identify various abnormalities of locomotion dynamics. Moreover, it allows one to diagnose qualitatively and quantitatively serious brain and central nervous system diseases.