Purchase this article with an account.
Jamie D Roitman, Elizabeth M Brannon, Michael L Platt; Representation of numerical magnitude in posterior parietal cortex. Journal of Vision 2006;6(6):690. doi: 10.1167/6.6.690.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Posterior parietal cortex has been implicated in the representation of number in humans and monkeys. Recently, Nieder and Miller (2004) reported that single parietal neurons selectively responded to visual arrays containing 1–5 elements, and that tuning bandwidth was proportional to preferred magnitude. However, spatial selectivity of the neurons was not considered, and the proportion of neurons sensitive to number was small (∼20%). To address these issues, we used an implicit numerical discrimination task to measure responses in the lateral intraparietal area (LIP) to visual arrays comprised of 2–32 elements. On each trial, a numerical cue predicting reward size was presented briefly while monkeys planned a saccade to a remote target. Stimulus controls balanced density, total pixels, and cumulative circumference. In each block, a standard numerosity (2,4,8,16 or 32) was presented on 50–60% of trials predicting a small reward, while deviant numbers were presented on the remaining trials predicting a large reward. For each neuron studied, the response field (RF) was first mapped, then the monkey performed 2–4 blocks of the implicit discrimination task with the numerical cue located in the RF and the saccade target in the opposite hemifield. In contrast with Nieder and Miller's (2004) report, we found more than half of LIP neurons were sensitive to the number of elements in the visual array and, moreover, neuronal activity varied monotonically with numerosity. Activity did not depend on stimulus frequency or reward expectation, suggesting that some parietal neurons encode the numerical magnitude of visual elements within their response fields.
This PDF is available to Subscribers Only