Abstract
Purpose: To better understand the mechanisms of visual attention, the spatial distribution of two effects of attention (contrast gain and baseline change) were studied with fMRI.
Method: Eight subjects experienced two different conditions: simply fixating the center without a task (passive-viewing/P condition) or performing a two-back memory task on the color of the fixation (two-back/T condition). A full-field flashing checkerboard was presented as the background for both conditions. During both conditions, the color of the fixation changed from black to either red or green for 500 ms, every 2000ms. In the rest period, the fixation changed to blue. The flashing checkerboard's contrast was set as 0, 6.25, 12.5, 25, 50 or 100%. Both the contrast gain and the baseline change effects were estimated for each voxel by fitting a contrast response function (CRF): R=B+(Rmax*Cα)/(σα+ Cα), where R is the amplitude of the response, B is the baseline, Rmax is the asymptotic amplitude of the response, C is the contrast, α is the exponential term for the steepness and, σ is the semi-saturation contrast. The equation was fitted to the results for each voxel and the values of B and σ for the two conditions (Bt−Bp and σt−σp) compared. Bt−Bp [[lt]]0 indicates a suppression in baseline BOLD signal and σt−σp[[gt]]0 a decrease in contrast gain.
Results: The spatial distributions of the contrast gain and the baseline change effects differ from each other. The T condition, compared to the P condition, shows a decrease of contrast gain (σt−σp[[gt]]0) for all eccentricities; the effect is weaker in V1 than in extra-striate cortex. The baseline change effect, however, varied with eccentricity: Bt−Bp [[gt]] 0 in the center and Bt−Bp [[lt]] 0 in the periphery.
Conclusion: The different spatial distribution of the two effect of attention suggests that they take place through different mechanisms.
NIH/NEI Grant EY 02115 and The Dana Foundation