Fig. 14.

fMRI maps show high motion selectivity in humanV3A, but not in V3. A–D show the topography of motion-selective activity in lateral and superior occipital cortex in three representative hemispheres. The activation map (red through white) shows two main regions that responded significantly more to radially moving concentric rings, as compared with MR levels produced by viewing those same rings when they were stationary. The boundaries of areas V1,V2, V3, VP,V3A, and V4v are transposed from retinotopic field sign maps produced in the same hemispheres.A and B show the motion-selective map from the same hemisphere, produced by stimuli of two different contrasts. C and D show data similar to those in A from two additional hemispheres. A, C, and D are produced by a stimulus of ∼1.5% luminance contrast; in B the stimulus was ∼95% contrast. In all panels, human area MT/MSTd appears as an activated focus at the bottom right. The second main motion-selective focus, in superior occipital cortex (upper left of panel), lies within retinotopically defined area V3A. Note that motion-selective fMRI activity is not seen in humanV3. Comparing A and Bconfirms that the focus of maximal activity within MT/V5 andV3A remains essentially constant despite a large difference in stimulus contrast. However, additional motion-selective activity is recruited when high contrast is used. In Bthe recruited activity occurs at the retinotopic representation of the (peripheral, circular) stimulus border in V1—because this is the highest contrast border in the stimulus and becauseV1 responds poorly to low-contrast stimuli (Tootell et al., 1995a; Boynton et al., 1996). Such high-contrast stimuli also can activate V2 as well. The calibration bar equals 1 cm, on average.