RESUMO
This report aims to show that computer analysis of the electroencephalogram (EEG) can greatly enhance and accelerate its diagnostic capability, permitting rapid, on-line detection of changes in cerebral cortical activity. We describe the use of the computer-analysed EEG to detect critical cerebral under-perfusion due to lowered systemic blood flow during cardiac bypass surgery. Such hypo-perfusion frequently results in post-operative neurological deficits. We demonstrate that computer-generated maps of the scalp distribution of percent power in the frequency band can indicate areas of critical cerebral cortical ischaemia sufficiently rapidly to allow raising of perfusion rate and reversal of functional deterioration before the onset of permanent damage. We show that computer-analysed EEG can detect abnormal levels of band activity within seconds after a drastic fall in systemic perfusion, indicating the "probability of abnormality" by performing a z-transform of the measured levels, against age-adjusted norms stored in the computer. We then introduce data indicating that the computer-analysed EEG, by combining several z-transformed EEG measures, can detect both symptomatic and asymptomatic cerebrovascular disease more effectively than the 133 Xe regional cerebral blood flow technique. Since cerebrovascular disease is a major cause of adult mortality and morbidity in the West Indies, this technique should be of great interest in the present context (AU)
