Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Parcellation of the laryngeal motor cortex and its role for early voice processing in fetal and neonatal humans
Poster Session C, Thursday, October 1, 10:45 am - 12:45 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Alejandra M. Hüsser1, Besim Prenaj2, Michel Belyk3, Christopher J. Ritter1, Martina Noé4, András Jakab4, Alexis Hervais-Adelman1,5; 1University of Zurich, 2University of Bern, 3Edge Hill University, 4Children's Hospital Zurich, 5University of Geneva
The fetal auditory system first responds to environmental sounds beginning at around 24 weeks gestational age and substantially increases in sophistication through the third trimester of pregnancy. Abundant evidence exists that prenatal acoustic experience has post-partum consequences, manifesting as a preference for the maternal voice or enhanced neuronal responses to languages experienced in utero vs unfamiliar languages. There is additional evidence that newborn infants produce cries carrying the pitch accent of the language of the gestational environment, suggesting that in utero auditory experience drives not only perceptual learning, but also patterns of vocal production. Producing a pitch-accent in a cry requires laryngeal modulation, similarly to linguistic and emotional prosody. It has been hypothesized that the dual representation of the larynx in human motor cortex may be a critical feature of the system conferring human’s exceptional vocal motor control. In adults, a dorsal and ventral representation of the larynx have been reported (referred to as laryngeal motor cortex LMC), the dorsal LMC being particularly relevant for prosodic aspects of speech, while the ventral LMC is commonly related to speech fluency. This present study focuses on seeking the dual representation of the LMC in the fetal and infant developing brain. We used resting-state functional magnetic resonance imaging data from the Developing Human Connectome Project of 567 neonatal (447 term-born neonates) and 222 fetal samples from single pregnancies with normal brain development. Imaging data were acquired at 20-44 weeks postmenstrual age (PMA). We used k-means clustering to partition the left and right motor cortices according to whole-brain voxel-wise correlations (k=9 per hemisphere), in the subset of term-born neonates, (i.e., those delivered ≥ 37 post-menstrual age). We defined a set of fronto-temporal regions of interest (ROIs) comprising the classical cortical language network. The similarity between individual connectivity patterns and the reference was calculated in terms of Pearson correlation coefficients. Connectivity analysis in term-born neonates serves to establish a reference. Their ventral motor clusters show associations with primary auditory areas (r > 0.53), and areas commonly related to vocalisations (IFG, r > 0.60). Voxel-wise connectivity of motor cortex voxels and the selected ROIs show a certain gradient of non-neighboring similarity. Averaged R2 over ROIs between the reference group and subjects of all three groups, i.e., term-born and preterm born neonates and fetuses, show a gradual increase over gestational age. The associations between ventral motor areas and speech-related areas may represent a precursor of the mature laryngeal control network. While a dual representation remains to be identified, there are indices of a similar connectivity pattern in dorsal parts of the motor cortex. The stronger functional connections between the LMC and the articulatory brain network in term-born neonates compared to fetuses and preterm infants, suggests a gradual development of orofacial control relevant for refined articulation during the final weeks of gestation. The relevance of this development for newborn cries has not yet been investigated and remains to be addressed. Wide-ranging language deficits in preterm born children could however indicate an association with an immature articulatory brain network.
Topic Areas: Speech Motor Control, Language Development/Acquisition