Inter-seizure variability in thalamic recruitment and its implications for precision thalamic neuromodulation.

backgroundThalamic stimulation is a promising approach to controlling seizures in patients with intractable epilepsy. It does not, however, provide good control for everyone. A big issue is that the role of the thalamus in seizure organization and propagation is unclear. When using responsive stimulation devices, they must detect seizure activity before sending stimulation. So, it's important to know which parts of the thalamus are involved in different seizures.

methodsTo better choose thalamic targets for stimulation, we studied how different seizures spread to each stimulation target. Expert reviews and automated tools were used to identify seizure spread recorded from invasive recordings. We categorized seizures based on how they start and spread, and determined whether seizures reached thalamic areas early or late. We used generalized linear models (GLM) to evaluate which seizure properties are predictive of time of spread to the thalamus, testing effect significance using Wald tests.

resultsWe show that seizures with <2 Hz synchronized-spiking patterns do not spread early to the thalamus, while seizures starting with faster activity (<20 Hz) spread early to all thalamic areas. Most importantly, seizures that begin broadly across the brain quickly recruit the centromedian and pulvinar areas, suggesting these may be better stimulation targets in such cases. Alternatively, seizures that start deep in the temporal lobe tend to involve the anterior part of the thalamus, meaning the centromedian might not be the best choice for those seizures.

conclusionsOur results suggest that by analyzing electrical activity during seizures, we can better predict which parts of the thalamus are involved. This could lead to more effective stimulation treatments for people with epilepsy.