A network of topographic numerosity maps in human association cortex.

Ben M. Harvey, Serge O. Dumoulin

Research output: Contribution to journal/periodicalArticleScientificpeer-review

73 Citations (Scopus)
283 Downloads (Pure)


Sensory and motor cortices each contain multiple topographic
maps with the structure of sensory organs (such as the retina
or cochlea) mapped onto the cortical surface. These sensory
maps are hierarchically organized. For example, visual field
maps contain neurons that represent increasingly large parts of
visual space with increasingly complex responses1. Some visual
neurons respond to stimuli with a particular numerosity —
the number of objects in a set. We recently discovered a parietal
topographic numerosity map in which neural numerosity preferences
progress gradually across the cortical surface2, analogous
to sensory maps. Following this analogy, we hypothesized
that there may be multiple numerosity maps. Numerosity
perception is implicated in many cognitive functions, including
foraging3, multiple object tracking4, dividing attention5,
decision-making6 and mathematics7–9. Here we use ultra-highfield
(7 Tesla, 7T) functional magnetic resonance imaging
(fMRI) and neural-model-based analyses to reveal numerosityselective
neural populations organized into six widely separated
topographic maps in each hemisphere. Although we
describe subtle differences between these maps, their properties
are very similar, unlike in sensory map hierarchies. These
maps are found in areas implicated in object recognition, motion
perception, attention control, decision-making and mathematics.
Multiple numerosity maps may allow interactions
with these cognitive systems, suggesting a broad role for
quantity processing in supporting many perceptual and cognitive functions.
Original languageEnglish
Article number36
JournalNature Human Behaviour
Publication statusPublished - 2017


Dive into the research topics of 'A network of topographic numerosity maps in human association cortex.'. Together they form a unique fingerprint.

Cite this