CDN researchers Richard Rosch, Paul Hunter and Martin Meyer along with colleagues at UCL have recently published their findings applying dynamic causal modelling (DCM) to examine global brain activity in zebrafish during seizures and at rest. Their paper, published in PLoS Computational Biology, gives us hope of better understanding the abnormal neuronal dynamics during seizures in people with epilepsy.

Whilst our understanding of epilepsy and epileptic seizures is expanding through the development of new models, techniques and technology, detailed mapping of how localised aberrant activity is integrated across the brain during seizures is still missing. Linking this activity across such spatial scales has been challenging until now, in part because of the technical difficulty in measuring neuronal signatures concurrently at such a comparatively large scale.

In their innovative work, the researchers used DCM to combine computational modelling with imaging data. More technically, they used Bayesian model inversion to fit neuronal models to their empirical data combining widely-used neural mass models with Bayesian model inversion algorithms.

In larval zebrafish, Richard, Paul and their colleagues pharmacologically induced epileptic seizures and used light sheet imaging to look at activity in their brains. Using neuronal mass models, they then modelled changes in spontaneous neuronal activity during the seizure, allowing Bayesian inference on changes in effective network connectivity and their underlying synaptic dynamics. This revealed concurrent changes in synaptic coupling both in micro-circuits and across the whole brain.