Finally mapped: The brain region that distinguishes bits from bounty

This layout, referred to as a topographical map, is characteristic of all primary senses -- sight, hearing, touch, smell and taste -- and scientists have long assumed that numerosity, while not a primary sense (but perceived similarly to one), might be characterized by such a map, too.

But they have not been able to find it, which has caused some doubt in the field as to whether a map for numerosity exists.

Now, however, Utrecht University's Benjamin Harvey, along with his colleagues, have sussed out signals that illustrate the hypothesized numerosity map is real.

Numerosity, it is important to note, is distinct from symbolic numbers. "We use symbolic numbers to represent numerosity and other aspects of magnitude, but the symbol itself is only a representation," Harvey said. He went on to explain that numerosity selectivity in the brain is derived from visual processing of image features, where symbolic number selectivity is derived by recognizing the shapes of numerals, written words, and linguistic sounds that represent numbers. "This latter task relies on very different parts of the brain that specialize in written and spoken language."

Understanding whether the brain's processing of numerosity and symbolic numbers is related, as we might be tempted to think, is just one area that will be better informed by Harvey's new map.

To uncover it, he and his colleagues asked eight adult study participants to look at patterns of dots that varied in number over time, all the while analysing the neural response properties in a numerosity-linked part of their brain using high-field fMRI (functional magnetic resonance imaging). Use of this advanced neuroimaging method allowed them to scan the subjects for far fewer hours per sitting than would have been required with a less powerful scanning technology.

With the fMRI data that resulted, Harvey and his team used population receptive field modelling, which aims to measure neural response as directly and quantitatively as possible. "This was the key to our success," Harvey said. It allowed the researchers to model the human fMRI response properties they observed following results of recordings from macaque neurons, in which numerosity experiments had been conducted more extensively.

Their efforts revealed a topographical layout of numerosity in the human brain; the small quantities of dots the participants observed were encoded by neurons in one part of the brain, and the larger quantities, in another.

This finding demonstrates that topography can emerge not just for lower-level cognitive functions, like the primary senses, but for higher-level cognitive functions, too.

"We are very excited that association cortex can produce emergent topographic structures," Harvey said.

Because scientists know a great deal about topographical maps (and have the tools to probe them), the work of Harvey et al. may help scientists better analyse the neural computation underlying number processing.

"We believe this will lead to a much more complete understanding of humans' unique numerical and mathematical skills," Harvey said.

Having heard from others in the field about the difficulty associated with the hunt for a topographical map of numerosity, Harvey and colleagues were surprised to obtain the results they did.

They also found the variations between their subjects interesting.

"Every individual brain is a complex and very different system," Harvey explained. "I was very surprised then that the map we report is in such a consistent location between our subjects, and that numerosity preferences always increased in the same direction along the cortex."

"On the other hand," he continued, "the extent of individual differences … is also striking." Harvey explained that understanding the consequences of these differences for their subjects' perception or task performance will require further study.

This work was supported by NWO Vidi grant #452.08.008.