Dynamic Analysis of Encephalic Activity
The application of nonlinear dynamical analysis techniques to electroencephalograms (EEGs) and evoked potentials (EPs) is described. The first method involves the extraction of features from state-space trajectories reconstructed from EEG segments recorded immediately prior to the presentation of a light flash. The feature vectors were used in a clustering procedure to produce classes of similar prestimulus EEG segments. Averaged evoked potentials were computed for each class and distinct differences between EPs were observed. In a second experiment, a neurophysiologically-based model of EEG generation was tested to determine its response to impulse-like inputs. It was found that this model produced EP-like activity. This suggests that the generators of the EP are located in the cortex.
Unable to display preview. Download preview PDF.
- Babloyantz, A. and Destexhe, A.: The Creutzfeld-Jakob disease in the hierarchy of chaotic attractors. In: From Chemical to Biological Orgaization, Markus, M., Müller, S. and Nicolis, G. (Eds.), Springer-Verlag, 1988.Google Scholar
- Brandt, M.E., and Jansen, B.H.: The Relationship between Prestimulus Alpha Amplitude and Visual Evoked Potential Amplitude, International Journal of Neuroscience, in press, 1991.Google Scholar
- Breeden, J.L. and Packard, N.H.: Technical Report CCSR-91-3. Center for Complex Systems Research, University of Illinois, Urbana-Champaign, 1991.Google Scholar
- Gong Y. and Haton, J.P.: Text-independent speaker recognition by trajectory space comparison. ICASSP-90, Albuquerque, April 1990.Google Scholar
- Zouridakis, G.: Simulation of EEG and VEP Activity, M.Sc.-thesis, University of Houston Houston, 1990.Google Scholar