Abstracts

Rationale: Epilepsy has traditionally been defined by its phenotype and genotype. But epilepsy is inherently a disease of brain dynamics, which remain poorly understood and difficult to quantify. We recently introduced a new method of analyzing seizures based upon their dynamotype, using basic principles from physics that have been successful in numerous real dynamical systems. We validated this model in a large cohort of human patients, which led to several predictions about how brain state changes over time.

Methods: We measured dynamotype in the tetanus toxin rat model of epileptogenesis, rat hippocampal slices with high K+/ low Mg2+, and a computational model of dynamotype testing ictal aborting stimulation across a vast range of parameters.

Results: Dynamotype in the tetanus toxin and brain slice models demonstrated predictable changes in brain state. These changes correspond to a migration between different areas of the dynamotype map. The dynamical model demonstrated seizures with a DC shift are prone to aborting stimulation over a wide range of parameters; however, seizures without a DC shift are extremely resistant to stimulation.

Conclusions: These results suggest that the brain state migrates over long time scales, which can be modulated by epileptogenesis and medications. They also suggest that this brain state affects the ability of a seizure to be aborted with electrical stimulation. These results motivate further work to quantify seizure dynamotype to predict the type of nearby seizures and design appropriate treatments for them.

Funding: NIH R01NS094399, Michigan Medicine (Robbins Family Research Fund and Lucas Family Research Fund)