Several other model studies describe the potential importance of network randomization regarding the vulnerability to seizures. Netoff and co-workers found in a hippocampal slice model that seizures could be induced by changing the proportion of local versus long-distance connections [36]. As the neural network configuration was transformed into a more random network, seizure-like behavior was more likely to arise. Srinivas and others observed hippocampal rat neurons in vitro which were injured with an exposure to glutamate [37]. The neural network became hypersynchronous and fired bursts at high frequency after this injury, which they interpreted as induced epileptic activity. The network properties showed that the clustering coefficient decreased after injury: the network became more random as epileptic activity developed. Percha and others found that in a model of mesial TLE, epileptogenesis is characterized by structural changes in the neural network topology and axonal sprouting [18]. They showed in a two-dimensional model that an abrupt transition from an unordered local state to an ordered state of global coherence occurs when the network configuration changes from a small-world network to a more random
