Computerized EEG in cardiac surgery

Abstract

EEG analysis is indicated in those cases in which cerebral ischemia may arise from procedures performed during cardiac surgery. The EEG is extremely sensitive to brain dysfunction [1] caused, for instance, by cerebral hypoperfusion or changing metabolism caused by different anaesthetics [2, 3]. Due to its sensitivity, the EEG provides general information concerning the condition of the patient involved. Observing the EEGs during surgery remains a hazardous task because of the enormous amount of EEG information generated from several cases simultaneously. It is obvious that automatic analysis is essential to interpret this information within the limited period of the surgical procedure. Automatic analysis of the EEG during cardiac surgery is focussed on three hallmarks. First is detection of brain dysfunction related to anaesthesiological and surgical procedures with the shortest possible delay so that its cause can be discovered and the efficacy of the course of action can be checked. Second, it is essential to have a built-in warning/alarm system to indicate when the margins of safety have been reached. Automatic analysis without such a system is useless and brings one no further than conventional strip chart recording. Such a warning system is based on the experience of the clinical neurophysiologist and can be considered an artificial intelligence system. The third is documentation of brain function in relationship to all other parameters being monitored; this brings up its capabilities for evaluating the impact of new anaesthesiological and surgical procedures on brain function in this fast developing field.