Journal Article

Can responses to basic non-numerical visual features explain neural numerosity responses?

Harvey, Ben M., Dumoulin, Serge O.

2017

Overview

Humans and many animals can distinguish between stimuli that differ in numerosity, the number of objects in a set. Human and macaque parietal lobes contain neurons that respond to changes in stimulus numerosity. However, basic non-numerical visual features can affect neural responses to and perception of numerosity, and visual features often co-vary with numerosity. Therefore, it is debated whether numerosity or co-varying low-level visual features underlie neural and behavioral responses to numerosity. To test the hypothesis that non-numerical visual features underlie neural numerosity responses in a human parietal numerosity map, we analyze responses to a group of numerosity stimulus configurations that have the same numerosity progression but vary considerably in their non-numerical visual features. Using ultra-high-field (7T) fMRI, we measure responses to these stimulus configurations in an area of posterior parietal cortex whose responses are believed to reflect numerosity-selective activity. We describe an fMRI analysis method to distinguish between alternative models of neural response functions, following a population receptive field (pRF) modeling approach. For each stimulus configuration, we first quantify the relationships between numerosity and several non-numerical visual features that have been proposed to underlie performance in numerosity discrimination tasks. We then determine how well responses to these non-numerical visual features predict the observed fMRI responses, and compare this to the predictions of responses to numerosity. We demonstrate that a numerosity response model predicts observed responses more accurately than models of responses to simple non-numerical visual features. As such, neural responses in cognitive processing need not reflect simpler properties of early sensory inputs. •We quantitatively test whether numerosity-tuned neural response reflect other visual features.•We quantify how non-numerical features change in common numerosity stimuli.•Parietal fMRI voxels respond to changes in stimulus numerosity.•Non-numerical stimulus feature changes predict these responses less well than numerosity.•Cognitive neural responses need not reflect low-level stimulus properties.

• Publisher: Elsevier Inc,Elsevier Limited
• Subject: Acuity / Adult / Age / Animal behavior / Animal models / Attention / Brain Mapping - methods / Cognition / Cognition - physiology / Cognitive ability / Computer programs / Cortex (parietal) / Debates / Experiments / Female / Functional magnetic resonance imaging / Handedness / Humans / Image processing / Macaca / Magnetic Resonance Imaging - methods / Male / Mathematical Concepts / Mathematical models / Neural coding / Neuroimaging / Neurons / Parietal Lobe - physiology / Prefrontal cortex / Research methodology / Scaling / Segmentation / Substantia grisea / Visual cortex / Visual discrimination / Visual perception / Visual Perception - physiology / Young Adult