Associating spontaneous with evoked activity in a neural mass model of cat visual cortex

Spontaneous activity of the brain at rest frequently has been considered as a mere backdrop to the salient activity evoked by external stimuli or tasks. However, the resting state of the brain consumes most of its energy budget, which suggests a far more important role. An intriguing hint comes from the spontaneous activity patterns in visual area 18, which were observed with voltage sensitive dye in anaesthetized cat by Kenet et al. in 2003 (Nature 425:954-956). These spontaneous patterns closely resembled those evoked by visual stimulation with oriented gratings, except that cortex appeared to cycle between different orientation maps. Moreover, spontaneous patterns similar to those evoked by horizontal and vertical gratings, orientations presumed to be of particular relevance for behaviour, occurred more often than those corresponding to oblique angles. We hypothesize that this kind of spontaneous activity develops largely autonomously, providing a dynamical reservoir of cortical states, which are then associated with visual stimuli through learning.
To test this hypothesis, we used a biologically inspired neural mass model to simulate a patch of visual cortex. Spontaneous transitions were induced by modest modifications of the neural connectivity, establishing a so-called stable heteroclinic channel. Significantly, the greater frequency of horizontal and vertical orientation maps emerged spontaneously. We then applied bar-shaped inputs to the model cortex and used simple Hebbian learning rules to modify the corresponding synaptic strengths. After unsupervised learning, different bar inputs reliably evoked their associated orientation state; whereas in the absence of input, the model cortex resumed its spontaneous cycling. We conclude that the experimentally observed similarities between spontaneous and evoked activity of cat visual cortex can be explained as the outcome of a learning process that associates external orientation stimuli with autonomous neural activity.