Three-channel electroencephalogram montage in neonatal seizure detection and quantification

Abstract

We compared three-channel montage with 21-channel neonatal minimal placement montage for the detection of neonatal seizures and quantification of seizure burden. Thirty-five neonatal electroencephalograms were retrospectively and blindly reviewed by two independent readers. Tracings were analyzed in the three-channel montage for seizure number, duration, and quantification of seizure burden before reanalysis with the full 21-channel neonatal minimal placement montage. Seizures were identified using standard definitions of electroencephalographic seizure. Sensitivity, specificity, and interrater reliability were calculated. The sensitivity and specificity of three-channel montage for detecting seizures > 10 seconds were 91% and 100% for reader 1, respectively, and 82% and 96% for reader 2, respectively. The interrater agreement for detection of seizures was excellent (κ = 0.86, 94% percent overall agreement). For quantification of seizure burden, strong, positive correlation existed between assessments by full montage and three-channel montage (for reader 1, r = 0.945, n = 11, P < 0.0001; for reader 2, r = 0.902, n = 11, P < 0.0001), and strong correlation existed between the readers for three-channel montage (r = 0.879, n = 11, P < 0.0001). Despite its limitations, three-channel montage is useful in the detection of neonatal seizures and quantification of seizure burden.