Purchase this article with an account.
M. Saenz, G. T. Buracas, G. M. Boynton; Feature-based attentional effects in early human visual cortex. Journal of Vision 2001;1(3):344. doi: https://doi.org/10.1167/1.3.344.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
We investigated a mechanism for the feature-based allocation of visual attention using complementary fMRI and psychophysical methods. In the fMRI study, subjects were presented with two circular patches of random moving dots placed to the left and right of fixation. The task-relevant side consisted of two overlapping fields of upward and downward moving dots. The task-irrelevant side consisted of a single field of dots moving in either direction. On the task side, subjects performed a speed discrimination task alternately on the upward and downward moving dots. We found that the fMRI response to the unchanging, task-irrelevant dots modulated with the switching of attention on the task side; the measured response in areas V1, V2, V3, V3A, and MT+ increased when subjects attended to the matching direction compared to the opposing direction. This modulation occurred without changing the stimulus, eye position, or spatial locus of attention and can be attributed only to a feature-based allocation of attention. Does this attentional modulation affect the saliency of the stimulus? To test this hypothesis, we measured behavioral data using stimuli similar to the fMRI experiment except that both sides consisted of overlapping fields of upward and downward moving dots. Subjects simultaneously performed a speed discrimination task on one field of dots from each side. As predicted, subjects performed significantly better when dividing attention between two fields moving in the same direction (both up or both down) than when dividing attention between different directions (one up and one down). Together, these results indicate that attention to a direction of motion increases the saliency of stimuli throughout the visual field that share that direction of motion. These results are consistent with a ‘feature similarity gain model’ of attention (Treue and Trujillo, 1999) in which attention to a particular feature increases the gain of sensory neurons tuned to that feature.
This PDF is available to Subscribers Only