A system to detect the onset of epileptic seizures

During prolonged EEG monitoring of epileptic patients, the continuous recording may be marked where seizures are likely to have taken place. Several methods of automatic seizure detection exist, but few can operate as an on-line seizure alert system. Proposed is a seizure detection system that can be used in real-time to alert medical staff to the onset of a patient seizure and hence improve clinical diagnosis. Proposed is a system based on the seizure probability of a section of EEG. Final operation features a user-tuneable threshold to exploit the trade-off between sensitivity and detection delay and an acceptable false detection rate.The system was designed using 307 hours of scalp EEG including a total of 56 seizures in 13 patients. Wavelet decomposition, feature extraction and data segmentation were employed to compute the a priori probabilities required for the Bayesian formulation used in training, testing and operation.Results based on the analysis of separate testing data (354 hours of scalp EEG including 74 seizures in 15 patients) show an average sensitivity of 70.5% and a false detection rate of 0.25/hr. This average sensitivity is based on the successful detection of 47 of the 74 seizures with a mean detection delay of 10.8s and a delay of 10 seconds or less in 31 of these (66% of detections).The system is considered to be preliminary and results are promising enough to encourage further work on probability-based seizure detection. The tuning mechanism was also seen to add value to the use of the system