Understanding real-world audition with computational fMRI

This article highlights two prominent aspects of current neuroimaging research on real-world audition. The first aspect considers the integration of fMRI with conventional and AI-based computational models, discussing how this integration is enriching our understanding of auditory cortical representation and processing of meaningful sounds. This includes the acoustic-to-semantic transformation underlying the recognition of natural sounds. The second aspect examines the investigation of the auditory cortical activation at a mesoscopic scale (<1mm), enabled by laminar fMRI – a technique that combines high-resolution fMRI acquisitions with advanced data analysis methods aimed at discerning neural activity across cortical layers. It is suggested that the combination of computational neuroimaging with laminar fMRI will allow us to track the flow of computational representations mediating sound processing within the auditory cortical network and the interactions of the auditory cortex with other brain regions.