Seizure Prediction in Epilepsy Patients

Purpose/Objective: Characterize rigorously the preictal period in epilepsy patients to improve the development of seizure prediction techniques. Background/Rationale: 30% of epilepsy patients are not well-controlled on medications and would benefit immensely from reliable seizure prediction. Methods/Methodology: Computational model consisting of in-silico Hodgkin-Huxley neurons arranged in a small-world topology using the Watts-Strogatz algorithm is used to generate synthetic electrocorticographic (ECoG) signals. ECoG data from 18 epilepsy patients is used to validate the model. Unsupervised machine learning is used with both patient and synthetic data to identify potential electrophysiologic biomarkers of the preictal period. Results/Findings: The model has shown states corresponding to interictal and ictal periods in synthetic ECoG signals. Validation against patient ECoG data is in progress. Conclusions: This research has the potential to rigorously characterize the preictal period, with the possibility of identifying electrophysiologic biomarkers of the preictal period. Interprofessional Implications: Success of this research project would provide insights into the neurobiology of ictogenesis and would assist neurologists and neurosurgeons in providing improved treatment options for patients with refractory epilepsy. References: Mormann, F., Andrzejak, R.G., Elger, C.E., and Lehnertz, K. (2007). Seizure prediction: the long and winding road. Brain 130:314–333. Mormann, F. and Andrzejak, R.G. (2016). Seizure prediction: making mileage on the long and winding road. Brain 139(Pt 6):1625-1627. Nemzer, L.R., Cravens, G.D., Worth, R.M., Motta, F., Placzek, A., Castro, V., and Lou, J.Q. (2021). Critical and ictal phases in simulated EEG signals on a small-world network. Front. Comput. Neurosci. 14:583350, doi: 10.3389/fncom.2020.583350.