Spatio-Temporal Modeling of Absence Epileptic Seizures Using Depth Recordings

In this study, we spatially and temporally analyze the absence epileptic seizures recorded from different layers of somatosensory cortex of a Genetic Absence Epileptic Rat from Strasbourg (GAERS). Synchronous appearance of spikes in different layers of somatosensory cortex during the seizures is the most important indication of the recorded data because the data were recorded locally. In fact, when one spike appears during the seizures, we can consider a spike matrix comprising the spikes recorded from different layers of somatosensory cortex. We describe these spike matrices by a spatio-temporal model. Then, we estimate the model parameters using a factor analysis method. Experimental results show that there are two factors which randomly combine with a background factor and generate the spike matrices of absence seizures. Moreover, it is shown that the spike matrices originate from the bottom and the top layers of somatosensory cortex. We also propose a validation framework to show the generality of the obtained spatio-temporal analysis in different absence seizures.