Abstracts

Rationale: A neural network simulation with realistic cortical architecture [1] has been performed to both investigate the resulting dynamics of the intrinsic bursting and the effects of external stimulation. A constant Poisson background activity is used in the simulation, and is distributed uniformly over the modeled cortex. The resulting bursting behaviour interestingly demonstrates the spontaneous onset and cessation of bursting epochs lasting for seconds. This model was treated with stimulation pulses from an external circular electrode, to investigate the phase dependence of any stimulation effects on the underlying intrinsic bursting activity. This is an effort to explore the stimulation studies of Motamedi et al. [2] in subdural grid implanted patients exhibiting afterdepolarizations during mapping procedures.Methods: The neural network model represents a 1.6 mm X 1.6 mm region of cortex and consists of seven neuron classes organized in three dimensional space by cortical layer, synaptic polarity, and electrophysiological properties. The 65,536 cells are single compartment and follow a modified Hodgkin-Huxley dynamics. The intercellular connections are governed by histological data and published models of local cortical wiring. External field stimulation occurs at the axon initial segments (AIS) according to a set threshold on the derivative of the electric field along the direction of the AIS. The simulations are performed on a 16-node distributed 32-bit processor system.Results: The bursting behaviour exhibited by these simulations demonstrates several seconds of spontaneous activity followed by spontaneous cessation and subsequent activity regeneration(Figure 1). Time-frequency evaluation of the bursting phases reveals a stable bursting frequency during the active phases. Stimulation pulses provided by the external electrode in the model produces cessation of activity as well, with the most effective pulses being delivered close to the depolarization peak in the derived mean field potential (Figure 2). Conclusions: These spontaneous epochs of bursting are similar to the Class I type of epileptic behaviour of Lopes da Silva et al. [3]. The bursting frequency is strongly controlled by the network connectivity, and because of the randomness inherit in the bursting onset, the ictal state can not be predicted in time. The stimulation effects demonstrate a very strong phase dependence, similar to the experimental studies of Motamedi et al. [2], which also showed greater efficacy if the stimulation pulse was delivered closer to the depolarization phase of the extracellular potential. [1] Anderson WS, et al.(2007)Biol Cyber, accepted May 2007. [2] Motamedi GK, et al.(2002) Epilepsia 43:836-846. [3] Lopes da Silva F, et al. (2003) Epilepsia 44(Suppl. 12):72-83. Supported by the Epilepsy Foundation, and NIH NS 38958.