Wrs. Webber et al., AN APPROACH TO SEIZURE DETECTION USING AN ARTIFICIAL NEURAL-NETWORK (ANN), Electroencephalography and clinical neurophysiology, 98(4), 1996, pp. 250-272
We have developed an EEG seizure detector based on an artificial neura
l network. The input layer of the ANN has 31 nodes quantifying the amp
litude, slope, curvature, rhythmicity, and frequency components of EEG
in a 2 sec epoch. The hidden layer has 30 nodes and the output layer
has 8 nodes representing various patterns of EEG activity (e.g. seizur
e, muscle, noise, normal). The value of the output node representing s
eizure activity is averaged over 3 consecutive epochs and a seizure is
declared when that average exceeds 0.65. Among 78 randomly selected f
iles from 50 patients not in the original training set, the detector d
eclared at least one seizure in 76% of 34 files containing seizures. I
t declared no seizures in 93% of 44 files not containing seizures. Fou
r false detections during 4.1 h of recording yielded a false detection
rate of 1.0/h. The detector can continuously process 40 channels of E
EG with a 33 MHz 486 CPU. Although this method is still in its early s
tages of development, our results represent proof of the principle tha
t ANN could be utilized to provide a practical approach for automatic,
on-line, seizure detection.