State-dependent patterns of spatiotemporal coupling in rat visual cortex

Oscillatory population activity is ubiquitous in the mammalian neocortex. Oscillations are present across many frequency bands, and are postulated to reflect the integration of information over neural populations [1]. Such oscillations can be evoked by sensory stimuli, and neocortical populations can exhibit distinct forms of resonance to the stimuli. For instance, when subjects are exposed to flickering light at certain frequencies, their EEG-waves are prone to entrain, i.e. phase lock, to the stimulation frequency [2]. First, we address the state-dependence of such entrainment using a rat-model of sleep [3]). Under urethane anesthesia, the electrocorticogram (ECoG) alternates spontaneously between a low-voltage desynchronized state, which resembles the ECoG during rapid eye movement (REM) sleep, and a synchronized state, which resembles the ECoG during slow-wave sleep. This biphasic state represents an ideal model system to investigate state-dependent changes in the network responses to sensory stimuli. As expected, frequency coupling varied with both stimulus frequency (2.5 Hz to 15 Hz) and cortical state. Overall, the desynchronized state, indicative of a more activated cortex, is more susceptible to entrainment than the synchronized state, which is in agreement with previous reports (e.g. [4]). Furthermore, in some trials, the entrainment outlasts the stimulus train, giving way to poststimulus illusory responses resembling a dampened oscillation. Next, we use voltage-sensitive dye imaging (VSDI) with high signal-to-noise ratio [5] to study the spatiotemporal characteristics of single trials of this phenomenon, and describe the spatial coupling by cortical area and cortical state