Predicting Speech Disfluency in Five-Year-Olds using Gray Matter Volume

Abstract

Objective: The aim of this study was to investigate whether gray matter volumes in selected areas in the brain could serve as an indicator of speech disfluency in 5-year-old children. We identified regions of interest based on previous research and the DIVA model of speech production. The study included 120 participants (65M, 55F) and speech disfluency was measured as a continuous variable, covering a broad spectrum of speech disfluencies observed in a sample of children participating in the FinnBrain Birth Cohort study. Methods: T1 images obtained through MRI were preprocessed with the help of FreeSurfer software. The children's speech samples were recorded in both audio and video formats and were analyzed using the SALT software. Jansson-Verkasalo et al. 's (2020) and Yairi & Ambrose's (1999) classification system was used to categorize speech disfluencies into stuttering-like (SLD) and other (OD) types. The researchers employed a linear regression model to examine the correlation between gray matter volumes and speech disfluency for each hemisphere. Results: Regarding the left hemisphere, our examination revealed that disfluencies in children could be statistically predicted by the volume of gray matter located in the postcentral gyrus. We discovered that children with less gray matter volume in this region had higher levels of SLDs. In the case of the right hemisphere, it was observed that reduced gray matter volume (GMV) in the superior temporal gyrus (STG) and inferior parietal lobule (IPL) were significant indicators that participants were likely to experience more instances of speech disfluency. The study did not find any notable sex-related differences in the GMVs related to verbal disfluency. Conclusion: Stuttering theories based on the DIVA model have mainly focused on auditory maps, but our study suggests that the somatosensory cortex also plays a significant role in speech disfluencies. Our findings suggest that deficits in the somatosensory cortex may project speech disruptions during early development. Our discovery of decreased gray matter volume in the right STG is consistent with previous research, but our study differs in that we did not divide participants into stutterers and non-stutterers, which may explain why our associative findings diverge from previous studies. Further investigation is needed to confirm the validity of our results, especially in exploring the importance of gray matter volume in the right IPL in children with speech disfluency.

Abbreviations: DIVA = directions into velocities of articulators; SALT = systematic analysis of language transcripts