Figure 4.

Comparison with V5/MT Microstimulation in a Planar Depth Task

Psychometric data from this study were fitted with logistic regression [8] (red filled and unfilled circles) and compared to logistic regression fits to microstimulation data for a disparity discrimination task [11] (black filled and unfilled circles). The horizontal shifts were normalized by the threshold and plotted against the disparity-tuning index (DTI). While the mean DTI between the two studies did not differ (t test, p = 0.33), the normalized horizontal shift induced with microstimulation in our combined disparity-motion task (median 0.42 SD ± 0.63, n = 48) was larger than for DeAngelis et al.’s [11] disparity task (median 0.22 SD ± 0.24, n = 65; Wilcoxon rank sum, p < 0.01). Medians and SEM are shown in the bar graph. To exclude differences in site selectivity, we also compared only microstimulation sites with a significant shift. Again mean DTIs were comparable (t test, p = 0.56), but normalized shifts were considerably larger when microstimulation was applied during our task depending on the specific conjunction of disparity and motion (our data, median 0.84, SD ± 0.64, n = 32; DeAngelis et al., median 0.33, SD ± 0.22, n = 43; Wilcoxon rank sum p < 0.0001). p values in the figure key are for significant shifts induced by electrical stimulation (filled circles).