Effects of electrode deactivation on speech recognition in multichannel cochlear implant recipients

• Published in: Cochlear implants international Vol. 18; no. 6; pp. 324 - 334
• Main Authors: Schvartz-Leyzac, Kara C., Zwolan, Teresa A., Pfingst, Bryan E.
• Format: Journal Article
• Language: English
• Published: England Taylor & Francis 02.11.2017
• Subjects: Aged; Cochlear implant speech recognition; Cochlear Implantation - instrumentation; Cochlear Implantation - methods; Cochlear Implants; Electrodes, Implanted - utilization; Female; Hearing Loss, Sensorineural - physiopathology; Hearing Loss, Sensorineural - surgery; Humans; Male; Mapping; Middle Aged; Noise; Performance outcomes; Signal-To-Noise Ratio; Spectral resolution; Speech Perception; Mapping; Spectral resolution; Performance outcomes; Cochlear implant speech recognition
• ISSN: 1467-0100; 1754-7628
• DOI: 10.1080/14670100.2017.1359457

Objectives: The objective of the current study is to evaluate how speech recognition performance is affected by the number of active electrodes that are turned off in multichannel cochlear implants. Several recent studies have demonstrated positive effects of deactivating stimulation sites based on an objective measure in cochlear implant processing strategies. Previous studies using an analysis of variance have shown that, on average, cochlear implant listeners' performance does not improve beyond eight active electrodes. We hypothesized that using a generalized linear mixed model would allow for better examination of this question. Methods: Seven peri- and post-lingual adult cochlear implant users (eight ears) were tested on speech recognition tasks using experimental MAPs which contained either 8, 12, 16 or 20 active electrodes. Speech recognition tests included CUNY sentences in speech-shaped noise, TIMIT sentences in quiet as well as vowel (CVC) and consonant (CV) stimuli presented in quiet and in signal-to-noise ratios of 0 and +10 dB. Results: The speech recognition threshold in noise (dB SNR) significantly worsened by approximately 2 dB on average as the number of active electrodes was decreased from 20 to 8. Likewise, sentence recognition scores in quiet significantly decreased by an average of approximately 12%. Discussion/conclusion: Cochlear implant recipients can utilize and benefit from using more than eight spectral channels when listening to complex sentences or sentences in background noise. The results of the current study suggest a conservative approach for turning off stimulation sites is best when using site-selection procedures.