Stimulus-dependent activations and attention-related modulations in the auditory cortex: a meta-analysis of fMRI studies

Abstract

We meta-analyzed 115 functional magnetic resonance imaging (fMRI) studies reporting auditory-cortex (AC) coordinates for activations related to active and passive processing of pitch and spatial location of non-speech sounds, as well as to the active and passive speech and voice processing. We aimed at revealing any systematic differences between AC surface locations of these activations by statistically analyzing the activation loci using the open-source Matlab toolbox VAMCA (Visualization and Meta-analysis on Cortical Anatomy). AC activations associated with pitch processing (e.g., active or passive listening to tones with a varying vs. fixed pitch) had median loci in the middle superior temporal gyrus (STG), lateral to Heschl's gyrus. However, median loci of activations due to the processing of infrequent pitch changes in a tone stream were centered in the STG or planum temporale (PT), significantly posterior to the median loci for other types of pitch processing. Median loci of attention-related modulations due to focused attention to pitch (e.g., attending selectively to low or high tones delivered in concurrent sequences) were, in turn, centered in the STG or superior temporal sulcus (STS), posterior to median loci for passive pitch processing. Activations due to spatial processing were centered in the posterior STG or PT, significantly posterior to pitch processing loci (processing of infrequent pitch changes excluded). In the right-hemisphere AC, the median locus of spatial attention-related modulations was in the STS, significantly inferior to the median locus for passive spatial processing. Activations associated with speech processing and those associated with voice processing had indistinguishable median loci at the border of mid-STG and mid-STS. Median loci of attention-related modulations due to attention to speech were in the same mid-STG/STS region. Thus, while attention to the pitch or location of non-speech sounds seems to recruit AC areas less involved in passive pitch or location processing, focused attention to speech predominantly enhances activations in regions that already respond to human vocalizations during passive listening. This suggests that distinct attention mechanisms might be engaged by attention to speech and attention to more elemental auditory features such as tone pitch or location. This article is part of a Special Issue entitled Human Auditory Neuroimaging.

Keywords: 2D; 3D; AC; AG; ALE; BOLD; ERP; FM; HG; Heschl's gyrus; MEG; MMN; MNI; MTG; Montreal Neurological Institute; PET; PP; PT; SMG; STG; STS; VAMCA; Visualization and Meta-analysis on Cortical Anatomy; activation likelihood estimate; angular gyrus; auditory cortex; blood oxygenation level dependent; event-related brain potential; fMRI; frequency-modulated; functional magnetic resonance imaging; magnetoencephalography; middle temporal gyrus; mismatch negativity; planum polare; planum temporale; positron emission tomography; superior temporal gyrus; superior temporal sulcus; supramarginal gyrus; three-dimensional; two-dimensional.