TY - JOUR T1 - Thalamic activity patterns unfolding over multiple time scales predict seizure onset in absence epilepsy JF - bioRxiv DO - 10.1101/2020.03.04.976688 SP - 2020.03.04.976688 AU - Jordan Michael Sorokin AU - Alex Williams AU - Surya Ganguli AU - John Huguenard Y1 - 2020/01/01 UR - http://biorxiv.org/content/early/2020/03/05/2020.03.04.976688.abstract N2 - The brain has a remarkable, yet poorly understood, capacity to perform rapid dynamic switching between different cognitive states. Absence epilepsy, characterized by sudden transitions to and from highly synchronous thalamocortical oscillations, provides a unique window to investigate rapid state switching. Here we explored the transition into seizures in detail using simultaneous extracellular unit recordings from the thalamocortical circuit in the Scn8a mouse, a validated murine model of absence epilepsy. We find that trial-averaged neural firing in the thalamus, but not cortex, was transiently elevated several seconds prior to seizure onset. However, we observed large single-trial variability in pre-ictal dynamics both within and across subjects, suggesting possible heterogeneous transition dynamics into absence seizures. To quantify the single-trial amplitude and temporal variability, we developed a statistical model, which revealed that individual seizures are preceded by low dimensional neural dynamics that vary in amplitude and time across seizures. Interestingly, the single-trial pre-seizure amplitude modulation uncovered by the model showed strong periodicity over trials, suggesting that pre-ictal dynamics may co-modulate with arousal state. To our knowledge, our results are the first characterization of single-unit pre-ictal firing dynamics across the thalamocortical circuit in absence epilepsy. Our results argue that seizure-monitoring devices may be able to capitalize on seizure-by-seizure changes in pre-ictal activity to better predict seizure onset, and that the thalamus may be a source of clinically useful pre-ictal signatures. ER -