Epileptiform Activity Patterns in Coupled Neuronal Networks

We evaluate occurrence of seizure-like activity in computational network models based on the histology of mammalian cortex. In the simulations, seizure onset activity in a network (active patch) is connected to a larger patch of cells (passive patch). The following scenarios were investigated: (1) the activeand passive patches are contiguous with all connectivity intact; (2) active and passive populations are separated (≥1mm interdistance) with long-range connectivity dominating between them. We find that linking the active patch with a passive patch of sufficient excitatory susceptibility reduces pathological network bursting behavior otherwise present in the active patch, even over a separation of several mm. In addition, the balance of excitatory and inhibitory strength in the passive patch plays a crucial role in its ability to restrain network bursting in the active patch: unexpectedly, if the activity in the passive patch is too low, then the active patch can drive network bursting in both patches. Keywords—Network Model, Epilepsy, Parallel Computing