Support vector machines for seizure detection

The development of a robust technique for automatic detection of the epileptic seizures is an important goal in clinical neurosciences. In this paper, the support vector machines (SVM) have been used for this purpose. The system detects and uses the three features of the electroencephalogram (EEG), namely, energy, decay (damping) of the dominant frequency, and cyclostationarity of the signals. The different types of epileptic seizures have shown some common characteristics in the feature space that can be exploited in distinguishing them from the normal activity in the brain or the nonepileptic abnormalities. The use of SVMs achieves high sensitivity and at the same time shows an improvement in terms of computational speed in comparison with the other traditional systems.