Figure 2. Spatial attention does not change the location or shape of V1 receptive fields.

(A) Calcium imaging was done in L2/3 excitatory neurons expressing GCaMP6f.

(B) Structure of imaging sessions.

(C) Receptive fields (RFs) were mapped by reverse-correlating calcium responses to the sparse noise stimuli, which were shown during inter-trial intervals (ITI).

(D) Left, two simultaneously imaged fields-of-view in V1. Right, RFs of neurons in the two regions. The green and magenta rectangles covered regions that retinotopically corresponded to BR and TL locations, respectively. Red, ON field; blue, OFF field.

(E) Three example neurons with different RF locations. RFs were mapped separately in the three blocks in a session. Their geometric locations within the fields-of-view are shown in (D).

(F) A RF fitted with two-dimensional elliptical Gaussian.

(G) Between-block difference in distance from BR location to RF centroids. Black arrow, measured distance (zero at the center of BR location); r, radius of circular go/no-go stimuli. The area of go/no-go stimuli at BR location is shaded in green. n = 4,853 RFs.

(H) As in (G) for TL location. n = 5,307 RFs.

(I) Between-block difference in RF shape for RFs near and distant from BR location. n (≤ r) = 443 RFs, n (> r) = 4,414 RFs. p values were determined by hierarchical bootstrapping. Error bars, 90% CI of bootstrapped mean.

(J) As in (I) for TL location. n (≤ r) = 896 RFs, n (> r) = 4,414 RFs.

See also Figure S4.