Title :
Automated detection of epileptic seizures in the EEG
Author :
Hoeve, Maarten-Jan ; Jones, Richard D. ; Carroll, Grant J. ; Goelz, Hansjerg
Author_Institution :
Dept. of Med. Phys. & Bioeng., Christchurch Hosp., New Zealand
Abstract :
A system has been developed to detect epileptic seizures in real-time during long-term EEG (LTEEG) monitoring. LTEEG is an important clinical service provided by the Neurology Department at Christchurch Hospital to investigate patients who have relatively infrequent but recurring seizures over extended periods. The detection algorithm looks for extended amplitude and frequency changes, calculated using basic signal-processing techniques, followed by a rule-based stage which compares time-dependent features against dynamic thresholds for each channel. Spatial context is used to discriminate eye artifacts. The system was tested on EEG data from 5 patients containing 44 epileptic seizures. The sensitivity and selectivity of the algorithm were 88.7% and 92.6% respectively
Keywords :
computerised monitoring; diseases; electroencephalography; medical expert systems; medical signal detection; medical signal processing; patient monitoring; real-time systems; LTEEG; basic signal-processing techniques; clinical service; detection algorithm; dynamic thresholds; epileptic seizure automated detection; extended amplitude changes; extended frequency changes; extended periods; eye artifacts; long-term EEG monitoring; real-time; relatively infrequent recurring seizures; rule-based stage; selectivity; sensitivity; spatial context; time-dependent features; Biomedical engineering; Detection algorithms; Electroencephalography; Epilepsy; Frequency; Hospitals; Morphology; Nervous system; Patient monitoring; Physics;
Conference_Titel :
Engineering in Medicine and Biology Society, 2001. Proceedings of the 23rd Annual International Conference of the IEEE
Conference_Location :
Istanbul
Print_ISBN :
0-7803-7211-5
DOI :
10.1109/IEMBS.2001.1019104
