Processing of Numbers by Single Neurons in the Human Medial Temporal Lobe

Abstract

Decades of research have shown that animals and humans alike share an innate 'sense of number' that provides the cognitive start-up tool for the construction of all formal mathematical concepts. This system is anchored in a complex, highly distributed and interconnected neuronal 'number network'. In this thesis I could bridge the gap between single-unit recordings in animals and macroscopic functional imaging studies in humans, using the rare opportunity to record the activity of single neurons in the medial temporal lobe (MTL) of behaving human patients. In two different experimental protocols, calculation task and parity judgement task, we were able to explore several seemingly disparate aspects of numerical and mathematical cognition, addressing questions that had yet been eluded from investigation. As a first fundamental insight, we showed that single cells in the MTL can encode information about both quantities and simple arithmetic rules. These numerical representations follow a labelled-line coding. Segregated populations of neurons that encode numerosities and numerals with distinct tuning profiles, however, indicate different degrees of abstractness for nonsymbolic and symbolic stimulus formats. As a neuronal basis of numerical and arithmetic representations these cells may ultimately give rise to number theory and mathematics. Next, we revealed striking coding differences between small and large numerosities, mirroring subitizing and estimation processes that provide an intriguing link to the complex interplay of attention, working memory, and number representations. Finally, we uncovered static and dynamic coding mechanisms in different subregions of the MTL that do not only emphasize the MTL's role as an integral part of a wider cortical maths network, but equally important, highlight the substantial role this highly associative area also plays in working memory processes. All these findings provide valuable puzzle pieces that deepen our understanding of numerical representations constituting our 'sense of number'.