Abstracts

Rationale:
Understanding the relationship between individual neuron and network activityis a critical component in elucidating seizure mechanisms. The temporal relationship between multi-unit spiking activity and the low-frequency local field potential has been previously described to resemble a sinc function, or the unit impulse response of an ideal brick-wall filter. However, the spatial aspect of this relationship remains relatively uninvestigated. In this study, we characterize the spatiotemporal relationship between meso-scale spiking activity and the local field potential during a seizure by analyzing the ictal spike-triggered average (STA).
Method:
A theoretical-modeling approach was used to analyze clinical microelectrode array (MEA) from focal seizures. (1) A computational model of a network of neocortical pyramidal neurons was employed to produce a theoretical prediction for the spatiotemporal STA.(2) Ictal recordings were obtained from a microelectrode array implanted in a patient with focal epilepsy. Spike trains extracted from multiunit activity (300-3000 Hz) across the entire duration of seizures were used to create a spike-triggered average of the low-frequency component of the MEA local field potential (2-50 Hz) in both spatial and temporal domains.  Results(1) Model of a neocortical pyramidal cell population showed a spatial pattern of activity that approximates a sinc function.(2) This result was corroborated by observations in the patient MEA recordings. Spatiotemporal STAs of the low-frequency component of the MEA local field potential showed patterns resembling a sinc function. The peaks of the sinc function observed in the spatial domain were separated by a distance on the order of ~1mm, in line with known geometry of neocortical circuitry. Furthermore, the sinc functions observed in both the spatial and temporal domains allowed the relationship between these STAs to be linked by a complex exponential consistent with a Fourier transform.
Conclusion:
We examined human focal seizure activity and found that spatiotemporal activation of the cortex is governed by a Fourier-transform-pair relationship between time and location, which may be explained by local excitatory intracortical connectivity. This relationship indicates that middle-range (~1 mm) excitatory connections play a critical role in the activation patterns during the ongoing seizure. Unlike the close-range connectivity (< 1 mm), these connections are capable of promoting a pathological ‘escape’ of uncontrolled excitation in the cortex, thereby recruiting a critical mass of cortical neurons that are capable of sustaining the seizure. This work supports the development of novel therapies for pharmacoresistant epilepsy, such as stimulation approaches via surgical intervention to disrupt intracortical-grey matter connectivity in seizure propagation.   
Funding:
:R01 NS095368 R01 NS084142