Purchase this article with an account.
G. C. DeAngelis, T. Uka; MT neurons can account for behavioral performance in a depth discrimination task. Journal of Vision 2001;1(3):272. doi: 10.1167/1.3.272.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recent work has shown that cortical area MT contains a map of binocular disparities, and that electrical stimulation of MT neurons can bias judgements of depth. This suggests that MT neurons might be sufficiently sensitive to weak disparity signals to account for behavioral performance. To test this hypothesis, we recorded from isolated MT neurons in two rhesus monkeys during performance of a 2AFC depth discrimination task. Monkeys discriminated between two coarse disparities in random-dot stereograms, and task difficulty was titrated around psychophysical threshold by varying the percentage of binocularly correlated dots in the display. The visual stimulus was tailored to match the receptive field location and preferred velocity of each neuron. Behavioral thresholds (in % correlated dots) were measured at the 82% correct level from psychometric functions. Comparable neuronal thresholds were extracted from neurometric functions that were calculated for each single unit using ROC analysis. Across the entire sample of 104 neurons, the mean neuronal threshold (28.7%) was not significantly higher than the mean behavioral threshold (26.7%). Moreover, most neurons with strong disparity selectivity were more sensitive than the animal, and roughly 5 percent of neurons had thresholds lower than the monkey's best psychophysical threshold. We also computed “choice probabilities” to test whether trial-to-trial fluctuations in neuronal activity were predictive of behavioral choice. The average choice probability was 0.59 (0.62 and 0.56 for the two monkeys), which is larger than that reported previously for MT neurons in a direction discrimination task (Britten et al. 1996). Together, these findings indicate that monkeys could perform our depth discrimination task based on the activity of MT neurons.
This PDF is available to Subscribers Only