Detection of seizure patterns with multichannel amplitude-integrated EEG and the color density spectral array in the adult neurology intensive care unit

Abstract

This study's purpose was to determine the sensitivity, false-positive and false-negative of seizure detection in adult intensive care by amplitude-integrated electroencephalography (aEEG) and color density spectral array (CDSA).30 continuous electroencephalogram (EEG) recordings were randomly performed in 3 digital EEG-recording machines, 3 specialized neurophysiologists participated in this study, underwent 4 hours of training of CDSA and aEEG, marked any epochs suspected to be seizures without access to the raw EEG. The results will be compared and analyzed with continuous EEG reading to consider sensitivity, positive or negative rate.The recordings in this study, comprised 720 hours of EEG containing a total of 435 seizures. The median sensitivity for seizure identification was 80% of CDSA and 81.3% of aEEG, Median false-positive was 4 per 24 hours of CDSA, and 2 per 24 hours of aEEG display, Median false-negative was 4 per 24 hours of CDSA, and 4 per 24 hours of aEEG display. The time spent in identification of seizures by CDSA and aEEG was much time-saving than continuous EEG-reading.In this study, both CDSA and aEEG have a higher sensitivity but lower false-positive or missed rate in the interpretation of seizure identification in adult NICU.

Figure 1

Displays of seizures identified by CDSA and aEEG. The CDSA displays of seizures (A) and aEEG displays of seizures (B) were compared to seizures identified by analyzing the raw EEG data (X-axis on top of the displays), which were concealed from the neurophysiologists during the actual testing. A cluster of seizures in the raw EEG data is indicated by a black box on the X-axis, while the body-movement artifact is indicated by a gray box on the X-axis. aEEG = amplitude-integrated electroencephalography, CDSA = color density spectral array, EEG = electroencephalogram.