Stimulus evoked transient dynamics in the primary visual cortex in vivo : voltage sensitive dye and extracellular recordings

Abstract

In this study we investigated how the primary visual cortex responds to a change in the visual scene. In analogy to dynamical systems theory such temporal step input gives information about the rules governing the system.

One way to probe the dynamics of the visual system is to first display one image, A, and after a while replace that with another image, B. Despite that the new image B is stationary after the transition, the majority of the neurons will change their activity over time. Moreover the message that is being encoded by this activity may also change over time despite the fact that the new stimulus is constant.

Using voltage sensitive dyes and extracellular recordings this study examines the dynamics in the primary visual cortex in response to different types of stimulus transitions. The voltage sensitive dye is thought to mainly reflect sub-threshold dendritic membrane depolarization whereas extracellular recordings detect action potentials from a small population of neurons.

First we examined whether the onset of a luminance-defined square could evoke a correlate of feedback activity in the voltage sensitive dye signal. We found that the temporal maximum of the voltage sensitive dye signal in response to the onset of the square propagated from area 19 to area 17, i.e. opposite to a feed-forward propagation. This temporal maximum was reached at 90ms in area 19 and 100ms in area 17. The maximum propagated with a speed of 0.2-0.3mm/ms.

The next thing we studied was whether the depolarization at 100ms in areas 17 and 18 (90ms in higher areas) was correlated to any neuronal firing in area 17 and 18. To examine this we assumed that the temporal modulation of the instantaneous firing rate for each neuron could be described by a sum of multiple independent (across neurons) temporal components. We found that the typical initial transient at 50 ms could be described by two independent components and the later peak at 100 ms could be described with one component. The component that peaked at 100 ms was correlated to the voltage sensitive dye signal. Moreover the amplitude ratio between the 100ms component and the 50ms components was larger for the multiunits that represent the object background.

As the stimulus in papers I and II had relatively short duration, 83ms, the response to the stimulus offset could be superimposed on the 100 ms peak that was studied in those papers. To study this superposition we did a new set of experiments where we tested five different stimulus durations, 25, 50, 83, 133 and 250ms (paper III). Instead of an increase in the instantaneous firing rate and voltage sensitive dye after stimulus offset, as would be expected from a superposition of the off response and the existing activity, the activity decreased prior to the typical off-response for the shorter durations. For the shortest duration, even the on response was truncated by the stimulus offset.

Since both stimulus onset and offset generate a response (paper III), I examined the response when the onset of one stimulus is simultaneous with the offset of another stimulus, i.e. an image transition (paper IV). In response to an image transition from image A to B it was found that 1) the initial transient at 50 ms encodes the difference image (B-A) more strongly than the current retinal image (B), 2) whereas the later peak at 100 ms and onwards encodes the current retinal image more strongly than the difference image.

In summary, in response to a change of the visual scene the neuronal instantaneous firing rate reaches an absolute maximum at 50 ms after the scene change and a local maximum at 100 ms. This study shows that the peaks at 50 ms and 100 ms in the instantaneous firing rate separate in an independent component analysis. Those two peaks encode different retinal images despite that the retinal image is constant. Both peaks can be truncated by stimulus offset. Finally the peak at 100 ms coincided with a depolarization propagating from area 19 to area 17.