Spike Timing Dependent Plasticity Enhances Integrated Information at the EEG Level: A Large-scale Brain Simulation Experiment

The brain enhances functional segregation and functional integration through the developmental process. This is supported by neonatal electroencephalography (EEG) experiments using various measurements such as complexity dimension or integrated information. Elucidating the mechanisms of this simultaneous enhancement is important to understand how the brain becomes a sophisticated cognitive system. One candidate for this mechanism is the emergence of attractors because they are thought to represent differentiated information in a neural network and at the same time induce entrainment among multivariates. Indeed, in neuron level simulation studies, learning by spike timing dependent plasticity (STDP) rule resulted in the emergence of attractors. However, there is still a gap in knowledge between synaptic changes by STDP and brain activity at a larger spatial scale such as EEG. Thus, the main aim of this study was to test whether STDP learning affects functional segregation and integration at the EEG level. We constructed a large-scale brain simulation composed of spiking neurons and synapses mediated by STDP rule. We estimated EEG signals from simulated spiking neural activities. These novel systems enabled us to examine the effect of synaptic changes on EEG signals. As a measurement of well-balanced functional segregation and integration, we examined changes in integrated information $\phi^{\ast}$. We found the increases in $\phi^{\ast}$ and the number of attractors after STDP learning, and they showed a significant positive correlation. These results support our hypothesis that the emergence of attractors can be an underlying mechanism of enhanced functional segregation and integration.