Neural networks underlying magnitude perception: a specific meta-analysis of fMRI studies

Daily life requires simultaneously processing spatial, temporal, and numerical inputs to form a valid mental representation of the environment. The interrelation between these perceptions has been a subject of theoretical debate. For instance, a theory of magnitude (ATOM) asserts that magnitude perceptions are processed in overlapping brain areas, which has been tested in behavioral and neuroimaging studies. We aimed to combine functional magnetic resonance imaging (fMRI) results using a coordinate-based meta-analysis to test this primary assumption of ATOM regarding overlapping brain areas. We conducted separate literature searches for space, time, and number perception following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The analysis was based on 19 articles regarding space, 38 regarding time, and 31 regarding number perception. Coordinates were analyzed using the "Activation Likelihood Estimation" method, which focused on conjunction analysis. Double conjunction analyses revealed activations mainly in the fronto-parietal areas and insular cortex. The triple conjunction analysis revealed activations in the right hemisphere, specifically in the inferior parietal and inferior frontal areas (previously linked to magnitude perception) and the anterior insular cortex (implicated in interoception and salience). In support of the ATOM theory, these findings suggest that overlapping neural networks may underlie space, time, and number perceptions.