Real-time intraoperative monitoring during neurosurgical and neuroradiological procedures

A real-time intraoperative evoked potential (EP) monitoring system is described and evaluated. Unique features include (1) online artifact rejection to reduce noise contamination, (2) optimum digital filtering to improve the signal-to-noise ratio of the EP signal, (3) statistically defined confidence intervals to determine significant EP peak latency deviations, and (4) sliding windows of EP subaverages of various sizes to minimize feedback time to the surgeon. The reliability and validity of this system were determined by comparison with conventional intraoperative EP averaging and by examining the correlation of EP parameter changes with concurrent surgical and radiological manipulations. This system was clearly superior to conventional averaging systems. Reliable EPs could be obtained from neurologically compromised patients within the electrically hostile operating room environment, in cases in which conventional averaging failed to extract a stable EP signal. EP update times of 10-20 s were quite common and allowed direct moment-to-moment correlations with surgical and radiological events. Case histories are presented that show the utility of this system for aiding in the prevention of neurological complications. This utility is examined for neurosurgical and neuroradiological procedures involving spinal cord, brainstem, midbrain, and cortical structures, and affecting the somatosensory, motor, auditory, and visual system pathways.