engleski

Cognitive Evoked Potentials and Nonlinear Dinamics of Underlying EEG Activity

In the field of clinical neurophysiology the classical signal analysis approach is based upon the premises that the analysed system is linear, the process that generates the signal is stochastic and that the registered signals are stationary and ergodic. Because these conditions are only partly satisfied, at list in shorter segments of time, the use of classical linear signal analysis methods is possible. The new, modern approach assumes that the system is non-linear enabling the use of new mathematical tools, which in contrast to classical analytical approach enables the holistic one. These methods are based on the system states analysis trough observations of their stability, complexity and chaotic behaviour. In our work, we used both approaches in the analysis of EEG, sensory and cognitive evoked potentials and patient’ s reaction times measurement in the group of healthy subjects placed in different conditions, which assumes different grade of mental (cognitive) activity. The idea was to compare classical linear signal analysis methods with the non-linear approach in detecting the complexity of brain activities. Therefore, to enable the quantification of complexity of analysed mental activity, we used the homemade CognyPro stimulator. The patients EEG was recorded in following situations: Eyes closed, · ; Eyes opened, Eyes closed and patient performing count seven test, Eyes opened and patient performing count seven test, Patient passively observes the appearance of single numbers on the PC screen, Patients were stimulated with simple mathematical equations with correct and false results. They where asked to respond by pressing a button with their right hand when the result was correct. Patients were stimulated with simple mathematical equations with correct and false results. They where asked to use both hands: right one for correct and left for false results. The EEG was recorded during all conditions. Sensory and cognitive evoked potentials and patients reactions where recorded only when the stimulation was applied. The analysis results, including frequency analysis, autocorrelation function, correlation dimension, Lyapunov exponent and entropy, shows that the entropy and alpha band activity where best correlates of the intuitively expected complexity of brain electrical activity in described situations. The obtained results showed that the non-linear methods could be successfully applied in the field of clinical, behavioural and cognitive research.

eeg; evoked potentials; nonlinear dinamics

nije evidentirano