• epilepsy
• IEDs
• LIF
• local visual cortex
• spiking neural network

Epilepsy is one of the most common brain disorders, causing serious disability and premature death worldwide. The condition is primarily a disorder characterized by spontaneously occurring
seizures. This disorder is generally thought to be due to an unbalance between inhibition and excitation leading to runaway activity. An increase in excitation/inhibition drive ratio is a sufficient condition for the occurrence of epileptic spikes. Inhibition is mediated by GABA neurotransmitter which when released decrease activities of targeted neurons. For instance, other computational models confirmed that a progressive reduction of GABAergic inhibition might lead to the emergence of seizure-like episodes, characterized by the surge of an excessive and synchronous neural activity. The aim of our work is to study and characterize the transition toward epilepsy decreasing its GABAergic inhibition counterpart. We developed a randomly connected network of leaky integrate-and-fire neurons simulating local visual cortex activity in epilepsy, starting from previous works describing a balanced condition of the same network.
Progressively decreasing GABAergic conductances led the network to face two phase transition and to display epileptic activity. In line with other works, we found that epileptic spikes are generated by the synchronous discharges of a large number of pyramidal cells. On the other hand, the network displayed epileptic activity until the inhibitory neurons managed to silence the excitatory ones, ending therefore the ictal regime. We found that the frequency and the duration of intermittent epileptic episodes depended on (i) Strength of inhibitory synapses (ii) Size of the network (iii) External drive inputs. After that we will show how the Local Field Potential , which is the potential recorded in the extracellular space with an electrode, is affected by the emerging of epileptic activity.
In the first chapter we will introduce the basis of eletrical activity in the brain and how information is processed. We will show what are the main models used in epilepsy and focalise our attention on spiking neural networks. In the second chapter we will introduce Brian2, a flexible spiking neural network simulator and some tool of nonlinear temporal series analysis that will help us in the interpretation of the results. In the last chapter we will describe the characteristics of each phase of the network, we will characterize the transitions between them and finally we will see how Local Field Potential spectra is affected by this transitions.