Associations between semantic context in speech recognition, spatial release from masking, and executive functioning in children using bilateral cochlear implants and typical hearing children

Abstract

Speech materials that assess speech understanding in quiet and noise in children with and without hearing loss (HL) are typically highly predictable in their semantic context. Because natural environments contain unpredictable speech information, ecological validity is stronger with materials that manipulate semantic context. This study investigated the effects of semantic context on speech recognition and spatial release from masking in children (ages 9–15 years) with bilateral cochlear implants (BiCIs) and age-matched typical hearing (TH) children, and examined associations with executive functioning. Sentences were drawn from a corpus of semantically coherent or anomalous target sentences. Three conditions were tested: quiet, target and masker co-located, and target and masker spatially separated; speech interferers were presented at varying signal-to-noise ratios. TH children were also tested with vocoded stimuli simulating aspects of CI listening. Results revealed that children with BiCIs and TH may rely on both semantic and spatial cues differently to recognize degraded speech. Whereas speech recognition with clear speech was associated with cognitive flexibility and attention inhibition, auditory attention and inhibition were associated with understanding vocoded speech. Executive functioning of children using BiCIs may also be associated with their speech recognition in noise. [Work funded by NIH-NIDCD Grant R01DC020355 to R.Y. Litovsky.]