Quantitative electroencephalography versus automated detectors for seizure identification in the intensive care unit: Man v. machine
CCCF ePoster library. Lalgudi Ganesan S. Nov 7, 2018; 234658; 121
Dr. Saptharishi (Rishi) Lalgudi Ganesan
Dr. Saptharishi (Rishi) Lalgudi Ganesan
Login now to access Regular content available to all registered users.

You may also access this content "anytime, anywhere" with the Free MULTILEARNING App for iOS and Android
Abstract
Rate & Comment (0)

INTRODUCTION: Quantitative EEG (QEEG) tools such as amplitude-integrated EEG (aEEG) and color density spectral array (CDSA) can help EEG experts quickly and efficiently screen prolonged EEG recordings for seizures1-8. With most North American centers lacking round-the-clock neurophysiologist review9, automated seizure detection algorithms built within proprietary EEG monitoring systems could facilitate timely detection of seizures. However, there exist gaps in our understanding of how QEEG tools perform in comparison to the automated seizure detection algorithms.

OBJECTIVES: To compare the performance of commercially-available computerized seizure detection algorithms with QEEG-based seizure identification by EEG experts. 

METHODS: Nineteen EEG recordings, 11 of which contained 379 seizures, recorded from a representative cohort of critically ill children, were reviewed by a board-certified neurophysiologist who marked the beginning and end of all seizures as per standard criteria. We processed these 19 recordings using the default settings of four proprietary automated seizure detection algorithms: ICTA-S, (Newborn) NB, Persyst 11 (P11), and Persyst13 (P13). These recordings were then transformed to aEEG and CDSA. Three neurophysiologists and three EEG technologists independently marked all ‘suspected’ seizures on these QEEG trends using a digital review software. We calculated sensitivity for seizure identification and the false-positive rate ( FPR per 24 h) for each modality in comparison with those detected by the board-certified neurophysiologist using raw EEG. Sensitivity and FPR were calculated on a per-recording basis and on a per-seizure basis for focal, hemispheric and bilateral seizures, and for short (<3 minutes) and long (≥3 minutes) seizures.

RESULTS: For the automated detectors, median (range) sensitivity was 33.3% (0 – 100) with ICTA-S, 100% (52.6 – 100) with NB, 74.1% (37.5 – 100) with P11, and 91.2% (0-100) with P13. Neurophysiologists and EEG Technologists had a median (range) sensitivity of 84.6% (47.4 – 100) with CDSA and 82.4% (28.9 – 100) with aEEG for seizure identification (figure 1 & 2). Sensitivity was greater for generalized and long seizures in comparison to the hemispheric, focal or short seizures with all modalities (figure 3). Median (range) hourly false-positive rates were 0.04 (0.0 – 0.4) with ICTA-S, 4.8(0.2 – 11.5) with NB, 0 (0.0 – 2.8) with P11, 0.03 (0.0-4.45) with P13, 0.03 (0.0 – 0.42) with CDSA, and 0.02 (0.0 – 0.26) with aEEG. False positive rate was comparable among recordings with (11) and without seizures (8) for all modalities. There was a good agreement between CDSA and aEEG as well as between P11 and P13 on a per-recording basis, however there was poor agreement between automated detectors (P11,P13) and the EEG experts using CDSA and aEEG.

CONCLUSIONS:  Sensitivity of newer automated seizure detectors is approaching that of EEG experts using aEEG and CDSA, however false positive rates are significantly higher. The P13, which is the most recently developed detector, and the NB, which is a newborn seizure detector, were the automated algorithms that performed the best.  This suggests that newer automated seizure detection algorithms could complement neurophysiologists’ use of QEEG and facilitate efficient review of raw EEG with reasonable sensitivity. Advances in automated detection algorithms incorporating dynamic optimization to personalize seizure detection, should be the focus of newer studies.


    This eLearning portal is powered by:
    This eLearning portal is powered by MULTIEPORTAL
Anonymous User Privacy Preferences

Strictly Necessary Cookies (Always Active)

MULTILEARNING platforms and tools hereinafter referred as “MLG SOFTWARE” are provided to you as pure educational platforms/services requiring cookies to operate. In the case of the MLG SOFTWARE, cookies are essential for the Platform to function properly for the provision of education. If these cookies are disabled, a large subset of the functionality provided by the Platform will either be unavailable or cease to work as expected. The MLG SOFTWARE do not capture non-essential activities such as menu items and listings you click on or pages viewed.


Performance Cookies

Performance cookies are used to analyse how visitors use a website in order to provide a better user experience.


Save Settings