Figure 8.

Role of CRF ambiguity in surround integration. A, Long-bar stimulus (top) and long-bar control stimulus (bottom). The long-bar control stimulus corresponded to the complement of the bar within the CRF and moved in the same two local directions (indicated by arrows). Neurons that did not respond to this stimulus were judged to be subject to the aperture problem for long-bar stimuli. Short-bar control stimuli were constructed similarly (see Materials and Methods). B, Role of aperture problem in surround integration. For neurons subject to the aperture problem for long-bar stimuli, the average CMI was 0.24. This value was significantly positive (indicated by asterisk) and significantly different from the average CMI of 0.037 for neurons that were not subject to the aperture problem for those stimuli (t test p < 0.05). In response to contour stimuli, the average CMIs for these two groups of neurons (light gray bars) were both significantly positive (indicated by asterisks) and were not significantly different from each other. Error bars indicate SE.