An approach to measuring language-specific brain function in developmental language disorder

Area of Science:

• Neuroscience
• Developmental Psychology
• Linguistics

Background:

• Characterizing brain function in developmental language disorder (DLD) is challenging due to individual variability in task performance.
• Existing evidence on brain function in DLD is limited, necessitating refined methodologies.
• Understanding DLD's neural underpinnings is crucial for informing theoretical and clinical models.

Purpose of the Study:

• To characterize language-specific brain function in individuals with DLD using an adaptive semantic matching paradigm during fMRI.
• To employ a Bayesian analytical approach suitable for small clinical samples.
• To address methodological limitations in previous DLD neuroimaging studies.

Main Methods:

• An adaptive semantic matching paradigm with word-similarity and symbol-string judgments during fMRI.
• An adaptive task difficulty adjustment (2-up-1-down staircase) to control for individual performance variability.
• Bayesian analyses, including model comparison and second-level generalized linear models, were applied to fMRI data from adolescents with DLD (n=5) and neurotypical controls (n=12).

Main Results:

• Group differences in brain activation were observed in right hemisphere frontal, temporal, and semantic language homologue networks.
• No reliable differences were found in left hemisphere language networks or in brain lateralization between groups.
• The adaptive paradigm successfully elicited language-specific function while controlling for task performance variability.

Conclusions:

• The study introduces an adaptive fMRI paradigm and a robust Bayesian analytical approach for small, heterogeneous clinical samples.
• These methodological advancements may help differentiate sources of variability in DLD research, such as task performance versus inherent DLD characteristics.
• Future research employing these methods could yield significant insights into the neurobiology of DLD.