Improving speech recognition in noise of children with cochlear implants: Contributions of binaural input and FM systems

Cochlear implantation is typically recommended as an option for children and adults with severe-to-profound hearing losses when traditional amplification is not beneficial. The use of a monaural cochlear implant (CI) often provides children substantial gains in speech recognition in quiet and noise as compared to their performance with hearing aids (HA). However, users of CIs continue to experience significant difficulty in speech recognition in noise as measured by speech-recognition tests and self assessments. The purpose of this study was to create a sensitive speech recognition in noise test for young children and to explore the enhancement of speech recognition in noise for children with CIs and frequency modulated (FM) systems as a function of monaural versus binaural auditory input and the type of information provided during binaural input (acoustic or electric). Experiment One, on 32 adults and 10 children with normal hearing, showed that phrases for the new speech-recognition test in noise were equally intelligible to each other and that the testing procedure reliably predicted 50% correct speech-in-noise thresholds. Experiment Two included 12 children with bilateral CIs and 10 children with a monaural CI and a HA on the non-implant ear. Speech recognition was evaluated in monaural and binaural conditions with and without the use of an FM system. There were two conditions with no-FM system: (1) monaural input with a CI and (2) binaural input with a CI or HA on the second side. Stimulation arrangements for the four FM system conditions included: (1) monaural input with FM to the first CI, (2) binaural input with FM to the first CI, (3) binaural input with FM to the second CI or HA, and (4) binaural input with FM to both sides. No significant differences were found between the two groups (bilateral or bimodal) across any of the conditions. However, statistical differences were found across the four FM system conditions. Speech-in-noise thresholds in the binaural condition with FM system input to the second side were significantly poorer than in all of the other FM conditions. In addition, the monaural condition with FM system input was significantly poorer than the binaural condition with FM system input to both sides. The findings indicate that children with CIs perform best with FM system input to the first CI or to both sides. Analyses of benefit across individuals revealed that 19 of the 22 participants achieved a clinically-significant increase in speech recognition in noise of at least 5 dB when the FM receiver was added to the first CI compared to listening with the first CI alone. In general, use of two FM receivers resulted in the best performance, but an FM system evaluation is necessary to determine the most optimal listening arrangement for a child.