Poster Presentation

Search Abstracts | Symposia | Slide Sessions | Poster Sessions

Phonological Processing: A Critical Review and Meta-Analysis of 150 Functional Imaging Studies

Poster Session B, Friday, September 12, 4:30 - 6:00 pm, Field House

Rasika Viswanathan¹, Leonardo Fernandino¹, Priyanka Shah-Basak¹, Sara Pillay¹, Lisa Conant¹, Jeffrey Binder¹; ¹Medical College of Wisconsin

Phonological brain processes include mechanisms for perceiving speech sounds, activating long-term representations of speech sounds for tasks like naming and silent reading, and maintaining and manipulating such representations in working memory as needed. The neural correlates of these distinct facets of phonological processing are not yet clear, and prior functional imaging meta-analyses of phonological processing have not attempted to distinguish them. We performed a search of PubMed, Medline, PsychInfo, NeuroQuery, and Neurosynth databases for the years 1980-2023 using inclusive search terms for phonological processes and combined this list with corpora from previous phonology-related meta-analyses. Studies were reviewed if the published abstract indicated use of fMRI or PET in healthy adult individuals. We performed a full review of 1496 papers, 150 of which passed all inclusion/exclusion criteria and reported activation foci for at least one phonological contrast of interest. Contrasts were grouped in 4 categories: Phoneme Perception contrasts comparing semantically meaningless speech sounds to a low-level auditory control with comparable task requirements (41 studies); Phonological Retrieval contrasts comparing tasks or stimuli emphasizing retrieval of phonological codes from visual input (letter strings or pictures) with controls for low-level sensory, motor, semantic, and executive processes (44 studies); Phonological Short-Term Memory contrasts emphasizing maintenance of semantically meaningless phonological information in short-term memory (23 studies); and Executive Load contrasts using letter or number n-back task comparisons (42 studies). Coordinates were coded in MNI space and analyzed with GingerALE 3.0. The resulting maps were thresholded using a cluster-forming threshold of p < 0.001 and cluster-level family-wise error correction at p < 0.05. Phoneme Perception contrasts activated bilateral but left-lateralized areas of the superior temporal gyrus (STG), extending from the lateral aspect of Heschl’s gyrus to the superior temporal sulcus (STS). Phonological Retrieval contrasts activated an exclusively left hemisphere network including posterior STG and STS, posterior planum temporale, supramarginal gyrus, precentral sulcus, and posterior inferior frontal gyrus. STG involvement in the Retrieval studies was mostly posterior to the areas involved in Perception, with a small region of overlap in the posterior left STS. Phonological Short-Term Memory contrasts activated a primarily motor network including left precentral gyrus, left putamen, left supplementary motor area, and bilateral anterior insula. Executive Load contrasts activated a bilateral and more dorsal network including anterior intraparietal sulcus, inferior frontal junction and middle frontal gyrus, anterior insula, and supplementary motor area. Retrieval, Short-Term Memory, and Executive Load activations overlapped near the inferior frontal junction. The results highlight distinct networks supporting qualitatively different aspects of phonological processing corresponding to speech sound perception, mental code retrieval, code maintenance, and manipulation of phonological codes in working memory. The limited overlap between Perception and Retrieval networks suggests largely distinct input (mid-STG) and output (posterior STG) phonological streams in the superior temporal cortex.

Topic Areas: Phonology,