The Optimized Pitch and Language (OPAL) strategy enhances pitch perception through coding of fundamental frequency (F0) amplitude modulation information in the stimulus envelope delivered to a cochlear implant. Previous research using a prototype of the strategy demonstrated significant benefits in musical pitch and lexical tone discrimination tasks with no degradation in speech recognition when compared with the clinical Advanced Combination Encoder (ACE) strategy in a small group of subjects. Based on those studies, a modified version of the strategy was implemented in the commercial Nucleus CP900 series processor. The aims of the present study were to establish whether the CP900 OPAL implementation continued to provide improved F0 pitch perception in a speech intonation task with no degradation to speech perception in quiet and noise, when compared with the clinical ACE strategy in a larger cohort of subjects. Further aims were to evaluate fitting procedures and subject acclimatization to the strategy after take-home experience.
Twenty experienced adult cochlear implant recipients were enrolled in the study. Two subjects withdrew during the study leaving 18 sets of data for analysis. A repeated-measures single-subject design with take-home experience was used to test for improved speech intonation perception using OPAL compared with ACE and for comparable performance between strategies for open-set word recognition in quiet at two presentation levels, sentence recognition in adaptive 4-talker babble noise, and speech intelligibility ratings. The stimulation rate employed for OPAL was 1200 pulses per second/channel which was higher than the default clinical rate of 900 pulses per second/channel used for ACE by all subjects in the present study. Two variations of the OPAL “F0 restore gain” (the gain applied to restore the loudness of modulated channels) were investigated: “custom” measured per subject and “default” which was the average of all subject custom gains.
A significant group mean benefit on the intonation test of 8.5% points was shown for OPAL compared with ACE. There was a significant period of adaptation to OPAL with significantly poorer sentence in noise scores acutely and after only 2 weeks of take-home experience. After 4 weeks of take-home experience, comparable word perception in quiet and sentence perception in noise for OPAL were obtained. Furthermore, there was good subject acceptability in the field with comparable speech intelligibility ratings between strategies. Results of the fitting procedure showed that OPAL did not require any additional steps compared with fitting of ACE. A default F0 restore gain provided comparable outcomes to a custom gain setting.
The CP900 OPAL implementation provided a significant benefit to perception of speech intonation when compared with ACE. Comparable speech perception (in quiet and noise) and subjective ratings of speech intelligibility between strategies were also achieved after a period of acclimatization. These outcomes are consistent with results of earlier studies using prototype versions of the strategy and reaffirm its potential for improvement of F0 pitch perception in speech while preserving coding of segmental speech information. Furthermore, the OPAL strategy can be programmed into subject’s processors using the same fitting procedures used for ACE thereby simplifying its adoption in clinical settings.
1The Hearing CRC, Victoria, Australia;
2Department of Audiology and Speech Pathology, University of Melbourne, Melbourne, Victoria, Australia;
3Cochlear Ltd. Australia, East Melbourne, Victoria, Australia; and
4Cochlear Ltd. Australia, Macquarie University, New South Wales, Australia.
ACKNOWLEDGMENTS: The Cooperative Research Centres (CRC) Programme supports industry-led end-user-driven research collaborations to address the major challenges facing Australia. In addition, the authors acknowledge the support of the Bionics Institute and the support it receives from the Victorian Government through its Operational Infrastructure Support Program. The authors also wish to thank the research volunteers that generously donated their time to partake in this study, and Dr Brett Swanson of Cochlear Ltd. for his comments on the research. The authors also acknowledge the support of Dr Adrienne Paterson of the HEARing CRC for management of ethical submissions, and Sylvia Tari and Alex Rousset from the Royal Victorian Eye and Ear Hospital (RVEEH) Cochlear Implant Clinic for assistance with recruitment. All authors except for A.A. contributed to study conception and design. P.D. and A.A. contributed to acquisition of data and statistical analysis. All authors except Y.Y. discussed the results and implications and contributed to writing of the manuscript.
This research was supported by the HEARing CRC, established under the Cooperative Research Centres (CRC) Programme.
The authors have no conflicts of interest to disclose.
Address for correspondence: Andrew Vandali, The Hearing CRC, 550 Swanston Street, Carlton, Victoria 3053, Australia. E-mail: firstname.lastname@example.org
Received December 6, 2017; accepted May 8, 2018.