Neural correlates of vowel perception and production

Résumé En

Introduction: Although speech perception and production were traditionally considered as independent mechanisms, recent findings - of motor regions involved in perception, auditory regions in production as well as sensorimotor mirror neurons - suggest a possible functional coupling of perception and action systems. However, few brain-imaging studies investigated both systems within the same participants and few explored vowels. The present functional magnetic resonance imaging (fMRI) study is the first to examine both vowel perception and production directly within the same subjects. Methods: 14 native french right-handed participants lied supine in a Bruker 3T imager using MRI-compatible microphone and headphones. In the first task, participants passively listened to 9 vowels (/i/, /y/, /u/, /e/, /ø/, /o/, /ɛ/, /œ/ and /ɔ/) previously recorded from their own voice. In a subsequent task, they produced the same vowels according to visual instructions. Vowel productions were recorded and acoustically analysed for performance evaluation. Each task consisted of three blocks of 10 minutes with vowels presented or produced in a pseudorandom order. In total, 180 functional images were acquired per task (18 per vowel plus 18 resting trials). Data was analysed using the SPM5 sofware. A random-effect group analysis was performed using repeated-measures ANOVA. Activations were estimated for the perception vs.baseline, production vs. baseline, main effect and conjunction contrasts (p < .01, FDR corrected, clusters superior to 100 voxels). Results: The perception task involved the superior temporal gyrus bilaterally and the opercular part of the left inferior frontal gyrus (Broca's area). The production task involved bilateral activations of the sensorimotor cortex, extending rostrally to the premotor cortex and caudally to the inferior parietal lobule, the supplementary motor area and the anterior cingulate gyrus, the insula and the rolandic operculum, the basal ganglia, the thalamus, the cerebellum, the cuneus and the precuneus and the superior temporal gyrus. The conjunction analysis between speaking and listening conditions showed a shared neural network with common activations observed in the left inferior frontal gyrus and in the superior temporal gyri. Comparisons between the two tasks revealed specific decreased responses within the posterior part of the superior temporal gyri and within the supramarginal gyri during vowel production. Discussion: Although previous studies have demonstrated activation of the ventral premotor cortex in passive speech perception, frontal activations in vowel perception were here largely confined to the inferior frontal gyrus. The absence of response in the ventral premotor cortex might be due to more simple articulatory structure of the stimuli and/or the perception of participant's own voice that would reduce mapping processes of perceived vowel onto internal motor representations of articulatory gestures. Nevertheless, by demonstrating decreased responses within the superior temporal and supramarginal gyri in production as well as common activations in Broca's area, these results provide evidence for both internal motor simulation in vowel perception and auditory/somatosensory feedback control mechanisms in vowel production, and strongly suggest a functional coupling of speech perception and action systems.