Dynamic networks differentiate the language ability of children with cochlear implants

Area of Science:

• Neuroscience
• Audiology
• Developmental Psychology

Background:

• Cochlear implantation (CI) aids language development in prelingually deafened children.
• A significant number of children with CI still face challenges in language and reading acquisition.
• Understanding the neural basis of these varied outcomes is crucial for optimizing CI effectiveness.

Purpose of the Study:

• To investigate the neural underpinnings of language and reading skills in children with cochlear implants (CI).
• To compare brain activity and connectivity in CI recipients with good versus poor language and reading abilities.
• To identify potential neural biomarkers for predicting CI outcomes.

Main Methods:

• Utilized high-density electroencephalography (EEG) in a resting-state condition.
• Employed dynamic imaging of coherent sources (DICS) and temporal partial directed coherence (TPDC) for source analysis and connectivity.
• Compared 50 children with CI (good/poor language) to 25 age- and gender-matched normal hearing (NH) children.

Main Results:

• CI groups exhibited higher coherence in alpha, beta, and gamma bands compared to NH children.
• Distinct cortical/subcortical source profiles and effective connectivity were observed between CI groups with good and poor language skills.
• A support vector machine (SVM) model accurately predicted language and reading scores using source and connectivity data.

Conclusions:

• Increased neural coherence in CI users suggests heightened oscillatory activity coupling.
• Divergent neural patterns and connectivity in CI children indicate compensatory adaptations impacting language development.
• Identified neural differences may serve as biomarkers for predicting CI success.