Seizure Diagnosis in Children based on the Electroencephalogram Modellind by Gaussian Process Model

Document Type : Original Article (s)

Authors

1 PhD Student, Department of Bioelectric and Biomedical Engineering, School of Advanced Medical Technology, Isfahan University of Medical Sciences, Isfahan, Iran

2 Associate Professor, The Medical Image and Signal Processing Research Center AND Department of Bioelectric and Biomedical Engineering, School of Advanced Medical Technology, Isfahan University of Medical Sciences, Isfahan, Iran

Abstract

Background: A seizure is the physical findings or changes in behavior occur after an episode of abnormal electrical activity in the brain. Seizures may interfere with cardiorespiratory function and with nutrition and may have detrimental long-term effects on cerebral development. Electroencephalogram (EEG) is essential in diagnosis and management of seizures. Automatic seizure detection is very important in clinical practice and has to be achieved by analyzing the EEG.Methods: For automatic seizure detection, we used Gaussian process (GP) model and train it on the EEG signals recorded from some children between the ages of 1.5 to 16 years. After modeling EEG signal by GP model, two measures of output signal were derived: the variance of the predicted signal and the hyperparameter ratio. It was based on the hypotheses that because the EEG signal during seizure events is more deterministic and rhythmic, we can use the changing of these two criteria for seizure detection.Findings: During seizure events, the variance of the model output signal reduced and the hayperparameter ratio increased. The second measure was less successful but it had other advantages like robustness to model order selection.Conclusion: The GP modeling is a good method for seizure detection. Important objectives are to perform this detection as quickly, efficiently and accurately as possible. In this method, decisions are made accurate and with negligible delay.

Keywords


  1. Thibeault-Eybalin MP, Lortie A, Carmant L. Neonatal seizures: do they damage the brain? Pediatr Neurol 2009; 40(3): 175-80.
  2. Temko A, Marnane L, Boylan G, Lightbody G. Adaptive modelling of background EEG for robust detection of neonatal seizures. Proceedings of the 2012 IEEE EMBS Conference on Biomedical Engineering and Sciences (IECBES); 2012 Dec 17-18; Langkawi, Malaysia.
  3. Deburchgraeve W, Cherian PJ, De Vos M, Swarte RM, Blok JH, Visser GH, et al. Automated neonatal spike train detection as part of a neonatal seizure detection system. Proceedings of the Noninvasive Functional Source Imaging of the Brain and Heart and the International Conference on Functional Biomedical Imaging; 2007 Oct 12-14; Hangzhou, China.
  4. Shellhaas RA, Clancy RR. Characterization of neonatal seizures by conventional EEG and single-channel EEG. Clin Neurophysiol 2007; 118(10): 2156-61.
  5. Motaghi S, Niknazar M, Sayyah M, Babapour V, Vosoughi Vahdat B, et al. Alterations of the electroencephalogram sub-bands amplitude during focal seizures in the pilocarpine model of epilepsy. Physiol Pharmacol 2012, 16(1), 11-20.
  6. Gotman J, Flanagan D, Rosenblatt B, Bye A, Mizrahi EM. Evaluation of an automatic seizure detection method for the newborn EEG. Electroencephalogr Clin Neurophysiol 1997; 103(3): 363-9.
  7. Gotman J. Automatic detection of seizures and spikes. J Clin Neurophysiol 1999; 16(2): 130-40.
  8. Cherian PJ, Deburchgraeve W, Swarte RM, De VM, Govaert P, Van HS, et al. Validation of a new automated neonatal seizure detection system: a clinician's perspective. Clin Neurophysiol 2011; 122(8): 1490-9.
  9. Han Y, Hsin YL, Harnod T, Liu W. Features and futures: seizure detection in partial epilepsies. Neurosurg Clin N Am 2011; 22(4): 507-18, vii.
  10. Yang Z, Zhao Q, Liu W. Improving spike separation using waveform derivatives. J Neural Eng 2009; 6(4): 046006.
  11. Chua KC, Chandran V, Acharya R, Lim CM. Automatic identification of epilepsy by HOS and power spectrum parameters using EEG signals: a comparative study. Conf Proc IEEE Eng Med Biol Soc 2008; 2008: 3824-7.
  12. Gotman J. A few thoughts on "What is a seizure?". Epilepsy Behav 2011; 22(Suppl 1): S2-S3.
  13. Yadav R, Swamy MN, Agarwal R. Model-based seizure detection for intracranial EEG recordings. IEEE Trans Biomed Eng 2012; 59(5): 1419-28.
  14. Rasmussen C. Evaluation of Gaussian processes and other methods for non-linear regression [PhD Thesis]. Toronto, ON: University of Toronto; 1996.
  15. Faul S, Gregorcic G, Boylan G, Marnane W, Lightbody G, Connolly S. Gaussian process modeling of EEG for the detection of neonatal seizures. IEEE Trans Biomed Eng 2007; 54(12): 2151-62.
  16. Shoeb AH. Application of machine learning to epileptic seizure onset detection and treatment [PhD Thesis]. Cambridge, MA; Massachusetts Institute of Technology; 2009.
  17. Rzempoluck EJ. Non-linear analysis of spike waveforms for classification of benign rolandic epilepsy of childhood [MSc Thesis]. Simon Fraser University; 1992.
  18. Rohrbacker N. Analysis of electroencephologram data using time-delay embeddings to reconstruct phase space. Dynamics at the Horsetooth 2009; 1: 1-11.