Effects of Spatial Attention on Macaque Primary Visual Cortex

Citation

Press, William Alan (1998) Effects of Spatial Attention on Macaque Primary Visual Cortex. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/jt9f-av51. https://resolver.caltech.edu/CaltechETD:etd-02012008-111851

Abstract

Does spatial attention affect neuronal responses in primary visual cortex? This question has been addressed in several previous studies, either with negative results or with modest positive results that do not rule out the possibility of experimental artifacts. The present study addressed three critical facets of this question: are responses in V1 affected by whether spatial attention is engaged, are they affected by where attention is directed, and does attention influence the modulatory effects of stimuli shown in the non-classical surround? Answering these questions requires establishing the following: whether V1 responses vary with attentional condition; whether any response changes are attributable to systematic offsets in eye position; and whether any responses changes are due to direct modulation of visually evoked responses or whether they are indirectly due to changes in baseline activity.

Responses from isolated single cells in V1 were recorded in two awake behaving monkeys. The monkeys were trained to perform a same-different orientation discrimination task while maintaining fixation. There were three attentional conditions, as determined by a cue: attending to the cell's classical receptive field (CRF); attending away from the CRF; and a passive condition, where no cue was shown and the animals had to maintain fixation throughout the trial's duration. In all three conditions, stimuli were presented both in the CRF and in the opposite hemifield. These stimuli were shown either alone, surrounded by parallel-oriented bars, or surrounded by orthogonal oriented bars.

59 cells were recorded under all three attentional conditions. 11/59 (19%) cells showed a significant effect of attentional condition on responses; no cells showed a significant effect of attentional condition on surround modulation. This was evident over the population of mean responses, as well, and could not be attributed to systematic biases in eye position. Compared to passive fixation, responses were suppressed, on average, by 7% when attention was engaged away from the CRF; compared to when attention was directed away from the CRF, moving attention to the CRF facilitated responses by 15%, back to passive fixation levels (similar results were obtained for the set of all 99 cells recorded with attending-away and attending-to conditions). Analyzing baseline activity showed that these response differences were modulations of the stimulus-evoked responses themselves and not of the baseline firing rates. This was confirmed by analyzing the time course of these attentional effects; modulations began about 80 ms after stimulus onset, 30-50 ms beyond the onset times for the responses. In addition, the response onset times were unchanged between attentional conditions. These results are discussed both in the context of previous studies that have investigated attentional modulation in V1 and in the context of computational models that attempt to describe the neurobiological underpinnings of spatial attention.

Thesis Files