1998•Minerva Access (University of Melbourne)Open access

An analysis of high rate speech processing strategies using the Nucleus 24 cochlear implant [Abstract]

Andrew E. Vandali, DB Grayden, Lesley A. Whitford, Kerrie L. Plant, Graeme M. Clark

Open full text 1 citations

Abstract

Speech comprehension for a group of five users of the Nucleus 24 Cochlear Implant system was explored at three rates of electrical stimulation, 250, 807, and 1615 pulses per second per channel. For the high stimulus rate the analysis frequency was the same as for the medium rate condition. The study investigated the effect of varying rate of stimulation when using the electrode selection technique of the SPEAK strategy. This has been undertaken using a repeated ABC experimental design to account for learning and minimize ordering effects. Speech perception was assessed using open-sets of CNC words in quiet and open-sets of CLTNY sentences at signal-to-noise ratios from +20 to 0dB. Closed-sets of 19 vowels and 24 consonants were also presented, in the H/V/D and A/C/A context. The recognition and perception of distinctive features were assessed across strategies and patients. Preliminary speech perception results have shown no statistically significant difference in performance between the low and medium stimulation rates. However, significantly poorer results were observed for the high rate condition for some tests. Individual differences may be explained by the effects of rate of stimulation on speech features.

Open-access reader

About this research paper

What this paper is about

Speech comprehension for a group of five users of the Nucleus 24 Cochlear Implant system was explored at three rates of electrical stimulation, 250, 807, and 1615 pulses per second per channel. For the high stimulus rate the analysis frequency was the same as for the medium rate condition. The study investigated the effect of varying rate of stimulation when using the electrode selection technique of the SPEAK strategy. This has been undertaken using a repeated ABC experimental design to account for learning and minimize ordering effects. Speech perception was assessed using open-sets of CNC words in quiet and open-sets of CLTNY sentences at signal-to-noise ratios from +20 to 0dB. Closed-sets of 19 vowels and 24 consonants were also presented, in the H/V/D and A/C/A context. The recognition and perception of distinctive features were assessed across strategies and patients. Preliminary speech perception results have shown no statistically significant difference in performance between the low and medium stimulation rates. However, significantly poorer results were observed for the high rate condition for some tests. Individual differences may be explained by the effects of rate of stimulation on speech features.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Speech comprehension for a group of five users of the Nucleus 24 Cochlear Implant system was explored at three rates of electrical stimulation, 250, 807, and 1615 pulses per second per channel. For the high stimulus rate the analysis frequency was the same as for the medium rate condition. The study investigated the effect of varying rate of stimulation when using the electrode selection technique of the SPEAK strategy. This has been undertaken using a repeated ABC experimental design to account for learning and minimize ordering effects. Speech perception was assessed using open-sets of CNC words in quiet and open-sets of CLTNY sentences at signal-to-noise ratios from +20 to 0dB. Closed-sets of 19 vowels and 24 consonants were also presented, in the H/V/D and A/C/A context. The recognition and perception of distinctive features were assessed across strategies and patients. Preliminary speech perception results have shown no statistically significant difference in performance between the low and medium stimulation rates. However, significantly poorer results were observed for the high rate condition for some tests. Individual differences may be explained by the effects of rate of stimulation on speech features.

Key concepts: Cochlear implant, Audiology, Speech processing, Speech perception, Speech recognition, Psychology, Computer science, Medicine

Related papers

Back to paper searchBrowse research topicsOriginal source
An analysis of high rate speech processing strategies using the Nucleus 24 cochlear implant [Abstract] — Research Paper | ScholarLens