Application of information theory techniques to the analysis of time series obtained from biological systems

Resumo

The main objective is to distinguish different dynamical modes in EEG and ECG signals using. Our aim is to contribute to identify the adequate techniques to study the dynamics of complex systems by applying them to processing biological signals. Regarding EEG signals, we intend to focus on two problems : i) to find the differences between the dynamics of two behavioral states (wakefulness and slow wave sleep, which are associated with a characteristic degree of synchronization of the EEG), comparing "stationary" moments of EEG which are typical of each case, and ii) to detect transients called sleep spindles (trains at 14 Hz originated by a neuronal oscillator in the reticular nucleus of thalamus). For these transients we aim to identify suitable methods to differentiate between spindles in normal subjects and spindles in patients with history of thalamic stoke. Regarding ECG, using a simultaneous recording of ECG and respiration, we intend to find adequate tools to differentiate between the variability of normal trace and that from diabetic patients. (AU)