Non-traditional interpretation of conventional EEG curve analyses

Abstract

The electroencephalographic (EEG) curve is a highly complex formation representing cerebral system activities. On the one hand it may resemble featureless oscillations in alpha activity, on the other hand it looks like a very irregular tangle of mostly quite random lines even when the proband or patient is in a waking state. Recent findings also indicate that the EEG signal can be studied as a complex numerical series using signal analysis in terms of chaodynamic processes and describe its outcome as fractals or attractors. Using spectral Gabor analysis (GA), local coherence (LCA) and amplitude analyses (AA) we report results with regard to clinical experience and atractor character of epileptic activity. We have developed a simple coefficient method showing the state of EEG synchronisation and, at the same time, also the actual state of integration or complexity of the systems in the brain. Amplitude analysis shows the fractal feature of alpha activity and attractor descriptions related to of epileptic activity. These results might be useful for detection of mental states related to levels of wakefulness or somnolence.