Imaging Predictors of Cochlear Implant Outcomes in Children with Congenital Hearing Loss

Early implantation of children with congenital sensorineural hearing loss is associated with large individual variations in speech and language outcomes. Traditional studies of pediatric cochlear implantation (CI) reveal that age at implant and residual hearing are correlated with language outcome. However, younger age at implant and more residual hearing alone do not explain the majority of outcome variance, nor can these factors be used to accurately predict language on the individual child level. This chapter provides a review of studies exploring language prediction based on preoperative brain imaging to enable neural-based prediction of language. Techniques used included structural, diffusion, functional, and arterial spin labeling magnetic resonance, positron emission tomography, functional near-infrared spectroscopy, and electroencephalography. Findings from structural and functional magnetic resonance imaging (MRI) studies suggest that the auditory regions such as the auditory radiation, axons that carry auditory information to the cortex, and higher-level processing areas in the auditory association and cognitive networks are predictive of speech outcomes. Imaging studies of the brain also suggest that the reorganization of sensory processing correlates with language outcome. These studies add to a growing body of evidence that variations and/or plasticity in neural structure and function influence auditory skills and language after implantation of children with no observable brain abnormalities on standard preoperative clinical imaging. More recent studies using machine learning hold promise to enable prediction of language on the individual level, which may lead to more effective early intervention for children at risk of having low language outcomes.