Computer modelling of epilepsy

Key Points Computer modelling of epilepsy is a branch of systems biology, a science that aims to combine the discoveries made by reductionist approaches into systems in order to understand how primary pathologies and secondary reactions interact to produce disease. Epilepsy, a dynamical disease of the brain, is well suited to study from the perspective of dynamical systems. Epilepsy is a complex set of syndromes with the commonality of recurrent seizures. Not only do the many individual epilepsy syndromes have different causes, but most epilepsies develop owing to the interaction of many causes at molecular, cellular, network and developmental levels, defying efforts to define simple cause-and-effect relations and suggesting the need for computer modelling. Knowledge discovery and data mining provides the substrate and support for dynamical modelling and allows the findings to be applied back to the research and clinical settings. The various dynamical modelling techniques that are used include stochastic models, low-dimensional (lumped) deterministic models and detailed neuronal network models. Computer models are applied across the range of epilepsy phenomenology, from the molecular to the clinical. At the patient level, Markov models have been used to assess patterns of remission and relapse in pediatric epilepsy. At the molecular level, deterministic models can predict alterations in cellular activity with ion-channel mutations. Many seizure models simulate activity at the network level. Some of these are lumped models, which use mean-field approximations to reduce the activity of many neurons to simple oscillators that are then coupled to produce complex activity patterns. Other models incorporate the details of neural activity and synaptic interactions, in order to reach down to the molecular level at which drug effects take place. Uncommonly among areas of neuroscience research, computer modelling is immediately accessible through downloads of established models. An intrinsically collaborative activity, the future of the endeavour lies in the cooperative efforts of clinicians, experimentalists and modellers. As a dynamical disorder, epilepsy is an attractive target for computer modelling. Here, Lytton provides an overview of the different types of computer model that have been used to describe epilepsy and shows how they can provide new insights into the disorder. Epilepsy is a complex set of disorders that can involve many areas of the cortex, as w