Comparison of quantitative EEG parameters from four different analysis techniques in evaluation of relationships between EEG and CBF in brain infarction.

Abstract Four computerized EEG analysis techniques (power spectral density using Fast Fourier Transform, period and amplitude analysis, normalized slope descriptors and a zeroline crossing variant) were compared with each other with regard to regional cerebral blood flow (rCBF) in 22 patients with unilateral supratentorial brain infarction. rCBF was measured by an intravenous injection of 133 Xenon isotope. The mean frequency of the EEG was determined by all the 4 computer techniques. The frequencies by all these methods showed almost significant or significant correlations with regional cerebral mean flow (rCBF m ) in the infarcted hemisphere, the highest correlation existing between the power spectral density frequency and rCBF m ( r = 0.60, P m , the highest one being between delta activity and rCBF m ( r = −0.75, P m , the first 3 parameters entered were all from the power spectral density analysis. Thus, the power spectral density parameters seem to include the main information to be extracted from the EEG by these quantification methods. The correlation coefficients between the EEG and regional fast compartmental flow (rCBF f ) were lower than those between the EEG and rCBF m . No significant correlation existed between the EEG and rCBF in the contralateral hemisphere of the infarction patients. The results suggest that in the quantification of the EEG in cerebral ischaemia the power spectral density is superior to the other analysis techniques evaluated in this study.