Predictive Processing during a Naturalistic Statistical Learning Task in ASD.

Area of Science:

• Neuroscience
• Developmental Psychology
• Linguistics

Background:

• Speech segmentation and language acquisition rely on detecting linguistic regularities.
• Neurocognitive mechanisms of speech segmentation in typical development and autism spectrum disorder (ASD) are not well understood.

Purpose of the Study:

• Investigate neural signals of statistical learning in children with ASD and typically developing (TD) peers.
• Examine brain responses in auditory and frontal regions during naturalistic speech processing.

Main Methods:

• Used magnetoencephalography (MEG) to track neural responses in 15 children with ASD and 14 TD children (ages 8-12).
• Analyzed evoked neural responses to syllable sequences in a statistical learning corpus across three passage repetitions.
• Focused on left primary auditory cortex, posterior superior temporal gyrus (pSTG), and inferior frontal gyrus (IFG).

Main Results:

• TD children showed a neural index of learning: increased neural response amplitude with increased syllable surprisal, emerging after repeated exposure.
• Children with ASD did not consistently exhibit this learning pattern across all analyzed brain regions.

Conclusions:

• Findings suggest differences in statistical learning of speech patterns in children with ASD compared to TD children.
• Hypotheses include potential bottom-up sensory deficits or top-down processing difficulties in children with ASD.