Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Early Left-Lateralization of Language Production in Preschoolers: An fNIRS Study
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
Pauline Maes1, Meghan Nyhof1, Meryem Yücel1, Helen Tager-Flusberg1; 1Boston University
Background: Recent fMRI work suggests that the language network is left-lateralized by four years of age, although language-related brain responses may remain weaker than in adults. Functional near-infrared spectroscopy (fNIRS), a wearable and movement-tolerant imaging method, enables the investigation of language processing and production in more socially integrated contexts and in children unlikely to tolerate scanners. The present study examined cortical responses during passive listening and prompted sentence repetition in preschoolers. Methods: Participants were 15 typically developing children, aged 2;11 to 6;10 years old (in months: 61.47±14.67, 35-82). During fNIRS recording, children completed a prompted sentence repetition task with two counterbalanced conditions: Listen and Repeat. In each trial, children viewed a short animation while an examiner produced a simple sentence describing the scene. In the Listen condition, children listened passively. In the Repeat condition, they repeated the sentence aloud. Trials were separated by a 10-second fixation cross. Hemodynamic responses were recorded using a NIRx NIRSport2 system with 16 sources and 16 detectors spaced at 30 mm. The probe covered bilateral anterior (inferior/middle frontal gyri) and posterior (temporo-parietal junction/superior temporal cortex) language regions. Data were preprocessed in Homer3 using a standard pipeline including channel pruning, motion correction, filtering, and conversion to hemoglobin concentration changes. Condition-specific hemodynamic responses were estimated from -2 to 14s using a generalized linear model. To account for developmental variability in head size and probe placement, analyses used a split-half functional channel-of-interest (fCOI) approach (Liu et al., 2022). For each participant and region of interest (left anterior, left posterior, right anterior, right posterior), the two most responsive channels to speech were identified independently across halves of the data and averaged. Mean HbO responses were extracted over the 4–12s window. One-sample t-tests assessed whether activation exceeded baseline within each condition and whether responses were greater during Repeat than Listen in each region of interest. Results: HbO concentration during the Listen condition did not significantly differ from baseline in any region of interest (all p>.3). During the Repeat condition, HbO concentration significantly exceeded baseline in the left anterior region only (t(14)=1.85, p=.04), with no significant difference in the remaining regions (all p>.3). Direct comparisons between conditions further showed greater HbO concentration during Repeat than Listen in the left anterior region (t(14)=3.58, p=.003), with no significant condition differences observed in the other regions of interest (all p>.2). Discussion: Preschoolers showed increased activation in left frontal language-related regions during repetition, consistent with early left-lateralization of language production in the frontal cortex. In contrast, passive listening did not elicit significant activation in targeted language regions. This may reflect immature specialization of temporal language networks in early childhood or methodological limitations, including imprecise cortical localization and stimuli that elicited neural responses too subtle to be reliably detected with fNIRS. Overall, these findings support the feasibility of using fNIRS to investigate language production in awake, freely moving preschoolers, opening opportunities to study clinical populations during a critical developmental window for language acquisition.
Topic Areas: Language Production, Language Development/Acquisition