This curve was then fitted with a Weibull function. The normalized neurometric curve showed a high similarity to the psychometric curve (r = 0.87 and 0.93 p < 0.01, for monkeys L and S, respectively). These results further support the notion that the population-response difference between circle and background can be useful for making a behavioral decision. Figure 5D displays the normalized population response as a function of orientation

jitter in the background area (left; monkey L; n = 9 recording sessions) and in the circle area (right; monkey S; n = 5 recording sessions). Epigenetic inhibitor The population response in the background is minimal for the contour condition (jitter = 0), and it increases with orientation jitter; i.e., the background suppression is decreasing with jitter (Figure 5D, left). The population response in the circle is maximal in the Smad inhibitor contour condition (jitter = 0), and it decreases with the orientation jitter; i.e., the enhancement in the circle is decreasing with the jitter (Figure 5D, right). Monkey L displayed a strong and significant negative correlation with the psychophysical performance in the background area (r = −0.74;

p = 0.02); however, the correlation in the circle area was small and positive but not significant (r = 0.14; p = 0.72). Monkey S displayed a strong positive and significant correlation with the psychometric curve in the circle area (r = 0.81; p = 0.03) but a nonsignificant negative correlation in the background area (r = −0.49; p = 0.32). These results can suggest that the monkeys were displaying ever different approaches of brain activity to process contour integration and then to segregate the contour from the noisy background. In other words, the monkeys may have used different weights for the circle and background areas in order to detect the contour from the noisy background. Although the correlation between contour saliency and neurometric curve is informative, the relation to the monkey’s perceptual report is still unclear. To study this, we compared

the FG-mjitt for orientation jitter trials, where the monkey was reporting either contour or noncontour with high probabilities. Because the stimulus remained the same and the report varied, it allowed us to test whether the observed modulations in V1 are linked to the monkeys’ perceptual report. Figure 6A displays the FG-mjitt as a function of time for two examples of orientation jitter conditions (±15 degrees in monkey L and ±10 degrees in monkey S). For both cases, the FG-mjitt in contour reported trials was higher in the late phase compared to the noncontour reported trials. This was true over multiple imaging sessions in both animals (Figure 6B. n = 6 and 9 orientation jitter conditions in which contour detection was 25%–75% in monkeys L and S, respectively; p < 0.05, paired sign ranked test).