Ictal and Post Ictal Impaired Consciousness due to Enhanced Mutual Information in Temporal Lobe Epilepsy

Seizure and synchronization are related to each other in complex manner. Altered synchrony has been implicated in loss of consciousness during partial seizures. However, the mechanism of altered consciousness following termination of seizures has not been studied well. In this work we used bivariate mutual information as a measure of synchronization to understand the neural correlate of altered consciousness during and after termination of mesial temporal lobe onset seizures. First, we have compared discrete bivariate mutual information (MI) measure with amplitude correlation (AC), phase synchronization (PS), nonlinear correlation and coherence, and established MI as a robust measure of synchronization. Next, we have extended MI to more than two signals by principal component method. The extended MI was applied on intracranial electroencephalogram (iEEG) before, during and after 23 temporal lobe seizures recorded from 11 patients. The analyses were carried out in delta, theta, alpha, beta and gamma bands. In 77% of the complex partial seizures MI was higher towards the seizure offset than in the first half of the seizure in the seizure onset zone (SOZ) channels in beta and gamma bands, whereas MI remained higher in the beginning or in the middle of the seizure than towards the offset across the least involved channels in the same bands. Synchronization seems built up outside the SOZ, gradually spread and culminated in SOZ and remained high beyond offset leading to impaired consciousness in 82% of the complex partial temporal lobe seizures. Consciousness impairment was scored according to a method previously applied to assess the same in patients with temporal lobe epilepsy during seizure. Keywords—Consciousness score, beta-band, gamma-band, intracranial EEG (iEEG), mutual information, temporal lobe seizure. Ictal and Post Ictal Impaired Consciousness due to Enhanced Mutual Information in Temporal Lobe Epilepsy Puneet Dheer*, Systems Science and Informatics Unit, Indian Statistical Institute, 8th Mile, Mysore Road, Bangalore 560059, India; e-mail: [email protected] Sandipan Pati*, UAB Epilepsy Center, Department of Neurology, University of Alabama at Birmingham, CIRC 312, 1719 6th Avenue South, Birmingham, AL 35294, USA; e-mail: [email protected] Srinath Jayachandran, Systems Science and Informatics Unit, Indian Statistical Institute, 8th Mile, Mysore Road, Bangalore 560059, India; e-mail: [email protected] Kaushik Kumar Majumdar, Systems Science and Informatics Unit, Indian Statistical Institute, 8th Mile, Mysore Road, Bangalore 560059, India; e-mail: [email protected] *shared first authors