Optical flow analysis to classify activity patterns in living neuronal networks

Abstract

Neuronal cultures provide an accessible way to observe and model the behavior of living neuronal networks. In this study, spontaneous activity recorded in such neuronal cultures is analyzed with the aim to find distinct spatiotemporal patterns of spontaneous activity, such as synchrony or spiral fronts. Human and rat neurons, cultured on flat or engineered surfaces, are analyzed along different days in vitro (DIV) using Neuropatt, a MATLAB toolbox designed for the detection of spatiotemporal patterns in neural population activity. Based on ‘optical flow analysis’, the toolbox reveals the emergence of characteristic patterns in the activity of the networks, and their dependence on DIVs and surface. Results indicate that development and surface properties change the connectivity of the network, which gives rise to macroscopic signatures such as characteristic patterns of activity