Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Predicting the auditory and language abilities of children with hearing loss using naturalistic stimuli-induced fNIRS signals
Poster Session B, Friday, October 25, 10:00 - 11:30 am, Great Hall 3 and 4
Yu Zhai1, Xiangyu He1, Kexin Jiang1, Chunming Lu1; 1State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Research, Beijing Normal University, China
Congenital hearing loss is a prevalent cognitive impairment affecting approximately 3 per 1000 newborns, resulting in severe deficits in auditory and language functions. While these children exhibit large variability in rehabilitation outcomes and struggle to achieve functional equivalence to their normal-hearing peers, it is crucial to have precise diagnosis and effective assessment of their auditory and language ability. Due to the challenges in measuring the brain functions from these children using traditional imaging techniques such as fMRI and EEG, there are no reliable neural markers established. To address these challenges, this study aimed to establish a novel approach in identifying the neural markers for children with hearing loss. In this study, fNIRS was used to measure brain functions from children under the age of six both with hearing loss (HL) and with normal hearing (NH), alongside their mothers, during a video watching task. The cortical regions covered the language processing system from low to high levels, including the bilateral sensorimotor and auditory cortex (SMAC), inferior parietal lobule (IPL) and prefrontal cortex (PFC). To systematically characterize the neural responses, we extracted three types of brain features: intrapersonal functional connectivity of children or mothers and the mother-child interpersonal neural synchronization (INS). Machine learning models were then employed to classify children between HL and NH and predict the auditory and language abilities of children with HL based on these brain features. Our findings revealed that for children with hearing loss, auditory and language abilities were more accurately predicted using either mother’s brain functional connectivity or mother-child INS compared to those derived from children’s brain signals. Moreover, mother-child INS also demonstrated better predictive capability for the improvement in abilities of hearing-impaired children six months later. Additionally, the brain markers for classifying HL vs. NH children were found to be dissociated from those for assessing the abilities of children with HL. Specifically, maternal functional connectivity of the lSMAC-lSMAC exhibited better classification of HL and NH children but could not predict HL children’s abilities, whereas the mother-child INS of the lSMAC-lIPL/rPFC and rSMAC-lSMAC displayed the opposite pattern. Therefore, maternal fNIRS signals induced by naturalistic stimuli could serve as reliable index to identify neural markers for classifying hearing deficits and assessing auditory and language abilities during recovery in children with HL. We proposed a “neural mirroring hypothesis”, suggesting that mothers’ brain function mirrors that of their children, offering a unique avenue for assessment when direct measurement of children’s abilities may be challenging. These findings not only provide promising clinical markers for personalized diagnosis and prognosis in hearing-impaired children but also highlight the importance of early mother-child social interaction in supporting children’s clinical rehabilitation. Keywords: Hearing loss children, Language development, Mother-child neural synchronization, fNIRS. Acknowledgements: This work was supported by the National Natural Science Foundation of China (62293550, 62293551, 61977008). Conflict of interest: The authors declare no competing financial interests.
Topic Areas: Language Development/Acquisition, Disorders: Developmental