Neonatal Brain Microstructure and Machine-Learning-Based Prediction of Early Language Development in Children Born
Rachel Vassar1, Kornél Schadl2, Katelyn Cahill-Rowley3
1Department of Orthopaedic Surgery, Stanford University School of Medicine, Stanford, California; Neonatal Neuroimaging Research Laboratory, Stanford University School of Medicine, Stanford, California; Division of Pediatric Neurology, Department of Neurology, University of California San Francisco, San Francisco, California.
Insights
Very-low-birth-weight preterm infants are at high risk for language impairments. Near-term brain imaging using MRI and DTI accurately identifies these infants, enabling early intervention for better language development.
Area of Science:
- Neuroscience
- Developmental Pediatrics
- Medical Imaging
Background:
- Very-low-birth-weight preterm infants exhibit higher rates of language impairments than full-term infants.
- Early identification of language delay risk in preterm infants is crucial for timely intervention during critical neuroplasticity periods.
- This study investigates the relationship between near-term brain structure (MRI) and white matter integrity (DTI) and early language development in very preterm children.
Purpose of the Study:
- To examine near-term structural brain MRI and diffusion tensor imaging (DTI) findings in relation to early language development in very preterm infants.
- To identify brain regions and white matter microstructural characteristics predictive of language outcomes.
- To develop accurate models for identifying preterm infants at high risk for language impairments.
Main Methods:
- 102 very-low-birth-weight neonates (birthweight ≤ 1500g, gestational age ≤ 32 weeks) were assessed.
- Near-term structural MRI evaluated white matter and cerebellar abnormalities.
- Diffusion tensor imaging (DTI) assessed white matter microstructure (fractional anisotropy, diffusivity measures).
- Language development was evaluated using the Bayley Scales of Infant-Toddler Development-III at 18-22 months adjusted age.
- Multivariate models with cross-validation identified predictive brain regions and logistic regression models predicted high-risk infants.
Main Results:
- Of 92 children tested, 31 scored below 85 on the composite language score, indicating moderate-to-severe delay.
- Cerebellar asymmetry was associated with lower receptive language subscores (P=0.016).
- DTI-based prediction models achieved high accuracy for identifying infants at risk for language impairments: composite (89% sensitivity, 86% specificity), expressive (100% sensitivity, 90% specificity), and receptive language (100% sensitivity, 90% specificity).
Conclusions:
- Near-term structural MRI and DTI-assessed white matter microstructure can aid in identifying very preterm infants at risk for language impairment.
- These neuroimaging techniques provide valuable tools for guiding early intervention strategies.
- Accurate prediction of language impairment risk allows for targeted support during optimal developmental periods.
Background:
Very-low-birth-weight preterm infants have a higher rate of language impairments compared with children born full term. Early identification of preterm infants at risk for language delay is essential to guide early intervention at the time of optimal neuroplasticity. This study examined near-term structural brain magnetic resonance imaging (MRI) and white matter microstructure assessed on diffusion tensor imaging (DTI) in relation to early language development in children born very preterm.
Methods:
A total of 102 very-low-birth-weight neonates (birthweight≤1500g, gestational age ≤32-weeks) were recruited to participate from 2010 to 2011. Near-term structural MRI was evaluated for white matter and cerebellar abnormalities. DTI fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity were assessed. Language development was assessed with Bayley Scales of Infant-Toddler Development-III at 18 to 22 months adjusted age. Multivariate models with leave-one-out cross-validation and exhaustive feature selection identified three brain regions most predictive of language function. Distinct logistic regression models predicted high-risk infants, defined by language scores >1 S.D. below average.
Results:
Of 102 children, 92 returned for neurodevelopmental testing. Composite language score mean ± S.D. was 89.0 ± 16.0; 31 of 92 children scored <85, including 15 of 92 scoring <70, suggesting moderate-to-severe delay. Children with cerebellar asymmetry had lower receptive language subscores (P = 0.016). Infants at high risk for language impairments were predicted based on regional white matter microstructure on DTI with high accuracy (sensitivity, specificity) for composite (89%, 86%), expressive (100%, 90%), and receptive language (100%, 90%).
Conclusions:
Multivariate models of near-term structural MRI and white matter microstructure on DTI may assist in identification of preterm infants at risk for language impairment, guiding early intervention.
