@misc{14894,
  author = {Tristan St{\"o}ber and Andrew Lehr and Ulrich Egert and Arvind Kumar},
  title = {On the origin of synchronous events in a network model of medial-temporal lobe epilepsy},
  abstract = {Medial temporal lobe epilepsy (MTLE) is associated with major structural changes in the dentate gyrus of the hippocampus. We focus on two prominent observations: (1) death of inhibitory and excitatory neurons in the hilus, notably inhibitory neurons are more prone to death in MTLE, (2) aberrant back-projections from and to granule cells (mossy fiber sprouting). We study their combined effect on network activity dynamics in a 2-dimensional network of excitatory and inhibitory neurons with local connections and conductance based synapses. Preferential death of inhibitory neurons at the lesion site and rewiring of lost connections leads to a ring of densely connected and overexcited neurons around the lesion site. In both the unlesioned and the lesioned network and over a wide range of external input strengths and relative levels of inhibition, spiking activity remains asynchronous irregular (AI). However, AI activity in the lesioned network is intermittently interrupted by transient synchronous network events (SNE). We show that SNEs persist even with strong global inhibition. Moreover, SNEs can be created by small perturbations far away from the lesion site. This suggests that even though SNEs arise due to the lesion, the trigger of SNEs could be anywhere in the network. We found that SNEs are preceded by synchronized sub-threshold membrane oscillations across the network. In such a state even a small perturbation is sufficient to synchronize the spiking activity and create a rapid wave of excitation across the network. In summary, we argue that SNEs reproduce several core features of epileptic seizures. Understanding the SNE mechanism may therefore contribute to the characterization and circumvention of seizure susceptibility in MTLE.},
  year = {2017},
  address = {NEST Conference 2017, J{\"u}lich, Germany},
}