ABSTRACT
Epilepsy is ranked the fourth most common neurological disorder after cerebral stroke, Alzheimer’s, and migraines. Most important applications of electroencephalography (EEG) recording and analysis is in epilepsy, for both clinical monitoring and research studies. For an epileptic brain, the EEG is extremely informative about the disorder during a seizure as well as in between two successive seizures. This chapter discusses the template-based automatic seizure detection algorithm proposed by Qu and Gotman on the continuously monitored EEG of epileptic patients. It was developed as a seizure warning system for clinical applications. The detection space contains a seizure template as a collection or a cluster of 5-dimensional vectors. There are precisely as many vectors in the seizure template cluster as there are epochs in the whole template. Time– frequency analyses are concerned about how frequency and amplitude of a signal are changing over time. Frequency is modulated by the up– down fluctuation of the amplitude crosspassing a predetermined horizontal baseline.
