10 Décembre – Thesis defense - Pierre Gilfriche

14 h Amphi Rougier (Pessac campus)

Nonlinear analysis of complex physiological control systems dynamics to assess performance and health in the connected human in an ecological environment.

LMaintaining a healthy state relies on many interactions between neural structures largely distributed in the organism and operating at different timescales. This structural complexity participates in the adaptiveness of physiological systems through a multi-level integration providing the control system with critical properties of flexibility, robustness and adaptability. This complexity in neurophysiological control systems can be evaluated through the analysis of their output signals: a complex signal is associated with a complex physiological network, and hence with an efficient control. Nonlinear analyses of physiological signals have proven their ability to characterize such complexity by introducing quantitative markers. These analyses bring new perspectives to the understanding of control strategies participating in health and performance in humans. Particularly, neurophysiological control distribution can be explored under different constraints to reveal adaptation strategies to numerous daily challenges. This way, studying the time course of heart rate, locomotor rhythms and postural sway is a way to explore control strategies contributing to health and performance in humans, in various contexts and under different constraints. In the present work, the frequency content, fractal fluctuations and entropy levels in time series generated by these systems are studied under different conditions. In a context of applied research, the experimental work has been conducted not only in the laboratory but also in ecological situations using connected devices (smartphone). For each situation, a methodological reflection is led to identify the most appropriate complexity analysis method for the control system studied, in line with the measurement devices used (connected devices or laboratory sensors).

Event localization