Structural gray matter features and behavioral preliterate skills predict future literacy - A machine learning...

Area of Science:

• Neuroscience
• Developmental Psychology
• Cognitive Science

Background:

• Children's neural systems adapt significantly for reading.
• Individual differences in brain structure may influence literacy and lead to disorders like dyslexia.
• Understanding these differences is crucial for early identification and intervention.

Purpose of the Study:

• To investigate interindividual differences in gray matter morphology before literacy acquisition.
• To determine if preliterate brain structure and cognitive-linguistic skills predict future reading ability.
• To identify specific neurostructural and cognitive predictors of literacy acquisition.

Main Methods:

• Longitudinal study of 42 German-speaking children from kindergarten to early school age.
• T1-weighted magnetic resonance imaging (MRI) for gray matter morphology assessment.
• Machine learning regression with leave-one-out cross-validation to predict literacy skills.

Main Results:

• Future literacy was predicted with high accuracy.
• Key predictors included gray matter volume in the left occipito-temporal cortex and local gyrification in specific left-hemisphere regions (insula, inferior frontal, supramarginal gyri).
• Phonological awareness was also a significant predictor of future literacy.

Conclusions:

• Preliterate brain morphology within the developing reading network can predict reading acquisition.
• Early neurostructural and cognitive-linguistic factors are vital for successful literacy development.
• This research offers insights into the neural basis of reading and potential markers for dyslexia risk.