Study of Feature Extraction Algorithms for Epileptic Seizure Prediction Based on SVM
Epilepsy is a common brain disease state, which threatens the safety of patients. So the effective prediction of epilepsy has great significance. To predict the epileptic seizure, energy feature of electroencephalogram (EEG) is extracted by wavelet transformation and power spectral. Then, support vector machine (SVM) is applied to separate the feature data. The research result shows that the energy of frequency band 0.5–8 Hz would rise 2000 s before seizure onset by analyzing inter-ictal and pre-ictal EEG’s wavelet energy. We used relative wavelet energy and SVM to analyze and test 8 patients’ EEG data, and it shows that the algorithm can predict some patients’ seizure onset except a few of patients’ bad behavior. We replace the wavelet with spectral power and use it to extract feature. The predict accuracy is improved by using spectral power and SVM. Comparing to the relative wavelet energy, the result of 6 patients’ test data improved by spectral power.
KeywordsEpileptic seizure prediction Wavelet energy Power spectral SVM
- Gbd 2015 Risk Factors Collaborators: Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet 388(10053), 1659–1724 (2015)Google Scholar
- Mallat, S.G.: A Theory for Multiresolution Signal Decomposition: The Wavelet Representation. IEEE Computer Society (1989)Google Scholar
- Shoeb, A.H.: Application of machine learning to epileptic seizure onset detection and treatment. Massachusetts Institute of Technology (2009)Google Scholar