How communication delay affects the collective behavior of neural networks

Because nothing occurs instantaneously in nature, time delay is a general feature of all interactions. Although the effects of delayed interaction are often neglected when the intrinsic dynamics is much slower than the coupling delay, they can be crucial otherwise. Synchronization, for example, is clearly affected by delays. A unique feature of delay-coupled oscillators is that their synchronization depends on the oscillatory frequency. We show that this feature can lead to transitions between synchrony and asynchrony in delayed coupled neuronal networks when the level of input to the network changes. The input changes the oscillatory period, which in the presence of delayed connections determines whether neurons fire in synchrony or incoherently. Moreover, since the impact of a neuronal network to downstream neurons increases when spikes are synchronous, networks with delayed connections can serve as gatekeeper layers mediating the firing transfer to other regions. The switch between active and inactive coupling channels depends on the input level, originating from sensory stimulus or from extrinsic control sources. This mechanism can regulate the opening and closing of communicating channels between cortical layers transmitting both firing rate as well as synchronous firing.