Lesley A. Whitford

Speech perception as a function of electrical stimulation rate: using the nucleus 24 cochlear implant system

Objective To investigate the effect of varying electrical stimulation rate on speech comprehension by cochlear implant users, while keeping the number of stimulated channels constant. Design Three average rates of electrical stimulation, 250, 807, and 1615 pulses per second per channel (pps/ch), were compared using a speech processing strategy that employed an electrode selection technique similar to that used in the Spectral Ma-ima Sound Processor strategy (McDermott, McKay, & Vandali, 1992; McDermott & Vandali, Reference Note 1; McKay, McDermott, Vandali, & Clark, 1991) and the Spectral Peak strategy (Skinner et al., 1994; Whitford et al., 1995). Speech perception tests with five users of the Nucleus 24 cochlear implant system were conducted over a 21-wk period. Subjects were given take-home experience with each rate condition. A repeated ABC evaluation protocol with alternating order was employed so as to account for learning effects and to minimize order effects. Perception of open-set monosyllabic words in quiet and open-set sentences at signal to noise ratios ranging from +20 to 0 dB, depending on the subject’s ability, were tested. A comparative performance questionnaire was also administered. Results No statistical differences in group performance between the 250 and 807 pps/ch rates were observed in any of the speech perception tests. However, significantly poorer group performance was observed for the 1615 pps/ch rate for some tests due predominantly to the results of one subject. Analysis of individual scores showed considerable variation across subjects. For some subjects, one or more of the three rate conditions evaluated provided benefits on some speech perception tasks. The results of the comparative performance questionnaire indicated a preference for the 250 and 807 pps/ch rates over the 1615 pps/ch rate for most listening situations. Conclusions For the speech processing strategy, implant system, and subjects evaluated in this study, the group results indicated that the use of electrical stimulation rates higher than 250 pps/ch (up to 1615 pps/ch) generally provided no significant improvement to speech comprehension. However, individual results indicated that perceptual benefits could be obtained by adjusting rate of stimulation optimally to suit each subject. Results from one subject, together with tinnitus problems arising from high-rate stimulation for another subject, indicated that high rates of stimulation may in fact be undesirable for some subjects.

Speech recognition with the Nucleus 24 SPEAK, ACE, and CIS speech coding strategies in newly implanted adults

Objective The objective of this study was to determine whether 1) the SPEAK, ACE or CIS speech coding strategy was associated with significantly better speech recognition for individual subjects implanted with the Nucleus CI24M internal device who used the SPrint™ speech processor, and 2) whether a subject’s preferred strategy for use in everyday life provided the best speech recognition. Design Twelve postlinguistically deaf, newly implanted adults participated. Initial preference for the three strategies was obtained with paired-comparison testing on the first day of implant stimulation with seven of eight U.S. subjects. During the first 12 wk, all subjects used each strategy alone for 4 wk to give them experience with the strategy and to identify preferred speech processor program parameters and settings that would be used in subsequent testing. For the next 6 wk, subjects used one strategy at a time for 2-wk intervals in the same order they had for the first 12 wk. At the end of each 2-wk interval, speech recognition testing was conducted with all three strategies. At the end of the 6 wk, all three strategies were placed on each subject’s processor, and subjects were asked to compare listening with these three programs in as many situations as possible for the next 2 wk. When they returned, subjects responded to a questionnaire asking about their preferred strategy and responded to two lists of medial consonants using each of the three strategies. The U.S. subjects also responded to two lists of medial vowels with the three strategies. Results Six of the 12 subjects in the present study had significantly higher CUNY sentence scores with the ACE strategy than with one or both of the other strategies; one of the 12 subjects had a significantly higher score with SPEAK than with ACE. In contrast, only two subjects had significantly higher CNC word and phoneme scores with one or two strategies than with the third strategy. One subject had a significantly higher vowel score with the SPEAK strategy than with the CIS strategy; and no subjects had significantly higher consonant scores with any strategy. Seven of 12 subjects preferred the ACE strategy, three preferred the SPEAK strategy, and two preferred the CIS strategy. Subjects’ responses on a questionnaire agreed closely with strategy preference from comparisons made in everyday life. There was a strong relation between the preferred strategy and scores on CUNY sentences but not for the other speech tests. For all subjects, except one, the preferred strategy was the one with the highest CUNY sentence score or was a strategy with a CUNY score not significantly lower than the highest score. Conclusions Despite differences in research design, there was remarkably close agreement in the pattern of group mean scores for the three strategies for CNC words and CUNY sentences in noise between the present study and the Conversion study (Arndt, Staller, Arcaroli, Hines, & Ebinger, Reference Note 1). In addition, essentially the same percentage of subjects preferred each strategy. For both studies, the strategy with which subjects had the highest score on the CUNY sentences in noise evaluation was strongly related to the preferred strategy; this relation was not strong for CNC words, CNC phonemes, vowels or consonants (Skinner, Arndt, & Staller, 2002). These results must be considered within the following context. For each strategy, programming parameters preferred for use in everyday life were determined before speech recognition was evaluated. In addition, implant recipients had experience listening with all three strategies in many situations in everyday life before choosing a preferred strategy. Finally, 11 of the 12 subjects strongly preferred one of the three strategies. Given the results and research design, it is recommended that clinicians fit each strategy sequentially starting with the ACE strategy so that the preferred programming parameters are determined for each strategy before recipients compare pairs of strategies. The goal is to provide the best opportunity for individuals to hear in everyday life within a clinically acceptable time period (e.g., 6 wk).

Pitch perception for different modes of stimulation using the Cochlear multiple-electrode prosthesis

Numerical estimations of pitch were obtained from nine postlinguistically deafened adults using the 22-electrode cochlear implant manufactured by Cochlear Pty. Limited. A series of electrodes on the array were stimulated using three modes of stimulation: Bipolar (BP), common ground (CG), and monopolar (MONO). In BP stimulation, an electric current was passed between two electrodes separated by one electrode for eight patients and two electrodes for one patient. In CG stimulation, a single electrode was activated and the other electrodes on the array were connected together to serve as the return path for the current. In MONO stimulation, an electric current was passed between a single electrode and the most basal electrode on the array. Pitch estimations were generally consistent with the tonotopic organization of the cochlea. There was a marked reversal in pitch for electrodes in the middle of the array using CG stimulation for three patients. A reduced range of pitch using MONO stimulation was recorded for patients where the most basal electrode was internal to the cochlea. There were also individual differences in pitch estimations between the three modes of stimulation for most patients. The current levels required to elicit threshold (T) and comfortable listening (C) levels were, in general, higher for BP stimulation than for CG stimulation and were lowest for MONO stimulation. For CG stimulation, there was a tendency for T and C levels to be higher for electrodes in the middle of the array than at the basal or apical ends. For MONO stimulation, T and C levels uniformly increased in an apical to basal direction for the majority of patients. There was no consistent pattern in T and C levels for BP stimulation. The size of the range of usable hearing using CG stimulation tended to be similar to that using BP stimulation and was usually higher than that using MONO stimulation.

Conversion from the SPEAK to the ACE strategy in children using the nucleus 24 cochlear implant system: speech perception and speech production outcomes.

Objective The main objective of this study was to assess whether speech perception and speech production in children using the Nucleus 24 cochlear implant system improved with a change in speech processing strategy from the SPEAK to the Advanced Combination Encoder (ACE) strategy. The major difference between the two strategies is that ACE uses a higher stimulation rate (in this study the stimulation rate was 900 Hz per channel) compared with the SPEAK strategy, where the stimulation rate is 250 Hz per channel. Information also was obtained regarding the adjustment period after conversion to the ACE strategy. Design An ABA experimental design was used where scores were initially obtained using the SPEAK strategy (in the initial A time interval), and subsequently performance was assessed using the ACE strategy (B time interval) and then again with the SPEAK strategy (second A time interval). The duration of the B interval was 10 wk, and the duration for the second A interval was 4 wk. Seven children aged between 9 and 16 yr who had at least 6 mo experience with the SPEAK strategy participated. Open-set monosyllabic CNC word perception in quiet and Speech Intelligibility Test sentence perception in noise was evaluated at the end of each of the time intervals. Word perception was also monitored at fortnightly intervals during the B time interval. Speech production was assessed at the end of the initial A time interval and at the end of the B time interval. Results Mean word and phoneme scores for open-set words in quiet for the group of seven children were significantly higher with the ACE strategy as compared with the SPEAK strategy scores obtained in both of the A time intervals. For sentences in noise, mean scores using the ACE strategy as well as the SPEAK strategy at the second A evaluation point were significantly higher than the scores using the SPEAK strategy measured at the first A time interval. This suggests that learning effects may have influenced outcomes. For some subjects, an initial decrease in scores was found during the initial 2-wk period after fitting the ACE strategy; however, scores subsequently were found to be similar to or higher than those when using the initial SPEAK strategy. Analysis of speech production assessments showed an improvement in the medial consonant scores after using the ACE strategy. Conclusions This study demonstrated that some children were able to benefit from the additional information provided by the ACE strategy as compared with the SPEAK strategy. However, the differences in overall performance between the two strategies appear to be relatively small.

Evaluation of the Nucleus Spectra 22 Processor and New Speech Processing Strategy (SPEAK) in Postlinguistically Deafened Adults

A new speech processing strategy (SPEAK) has been compared with the previous Multipeak (MPEAK) strategy in a study with 24 postlinguistically deafened adults. The results show that performance with the SPEAK coding strategy was significantly better for 58.31% of subjects on closed-set consonant identification, for 33.3% of subjects on closed-set vowel identification and open-set monosyllabic word recognition, and for 81.8% of subjects on open-set sentence recognition in quiet and in competing noise (+ 10 dB signal-to-noise ratio). By far the largest improvement observed was for sentence recognition in noise, with the mean score across subjects for the SPEAK strategy twice that obtained with MPEAK.

An investigation of input level range for the nucleus 24 cochlear implant system: speech perception performance, program preference, and loudness comfort ratings.

Objective Cochlear implant recipients often have limited access to lower level speech sounds. In this study we evaluated the effects of varying the input range characteristics of the Nucleus 24 cochlear implant system on recognition of vowels, consonants, and sentences in noise and on listening in everyday life. Design Twelve subjects participated in the study that was divided into two parts. In Part 1 subjects used speech processor (Nucleus 24 SPrint™) programs adjusted for three input sensitivity settings: a standard or default microphone sensitivity setting (MS 8), a setting that increased the input sensitivity by 10.5 dB (MS 15), and the same setting that increased input sensitivity but also incorporated the automatic sensitivity control (ASC; i.e., MS 15A) that is designed to reduce the loudness of noise. The default instantaneous input dynamic range (IIDR) of 30 dB was used in these programs (i.e., base level of 4; BL 4). Subjects were tested using each sensitivity program with vowels and consonants presented at very low to casual conversational levels of 40 dB SPL and 55 dB SPL, respectively. They were also tested with sentences presented at a raised level of 65 dB SPL in multi-talker babble at individually determined signal to noise ratios. In addition, subjects were given experience outside of the laboratory for several weeks. They were asked to complete a questionnaire where they compared the programs in different listening situations as well as the loudness of environmental sounds, and state the setting they preferred overall. In Part 2 of the study, subjects used two programs. The first program was their preferred sensitivity program from Part 1 that had an IIDR of 30 dB (BL 4). Seven subjects used MS 8 and four used MS 15, and one used the noise reduction program MS 15A. The second program used the same microphone sensitivity but had the IIDR extended by an additional 8 to 10 dB (BL 1/0). These two programs were evaluated similarly in the speech laboratory and with take-home experience as in Part 1. Results Part 1 Increasing the microphone input sensitivity by 10.5 dB (from MS 8 to MS 15) significantly improved the perception of vowels and consonants at 40 and 55 dB SPL. The group mean improvement in vowel scores was 25 percentage points at 40 dB SPL and 4 percentage points at 55 dB SPL. The group mean improvement for consonants was 23 percentage points at 40 dB SPL and 11 percentage points at 55 dB SPL. Increased input sensitivity did not significantly reduce the perception of sentences presented at 65 dB SPL in babble despite the fact that speech peaks were then within the compressed range above the SPrint processors automatic gain control (AGC) knee-point. Although there was a demonstrable advantage for perception of low-level speech with the higher input sensitivity (MS 15 and 15A), seven of the 12 subjects preferred MS 8, four preferred MS 15 or 15A, and one had no preference overall. Approximately half the subjects preferred MS 8 across the 18 listening situations, whereas an average of two subjects preferred MS 15 or 15A. The increased microphone sensitivity of MS 15 substantially increased the loudness of environmental sounds. However, use of the ASC noise reduction setting with MS 15 reduced the loudness of environmental sounds to equal or below that for MS 8. Results Part 2 The increased instantaneous input range gave some improvement (8 to 9 percentage points for the 40 dB SPL presentation level) in the perception of consonants. There was no statistically significant increase in vowel scores. Mean scores for sentences presented at 65 dB SPL in babble were significantly lower (5 percentage points) for the increased IIDR setting. Subjects had no preference for the increased IIDR over the default. The IIDR setting had no effect on the loudness of environmental sounds. Conclusions Given the fact that individuals differ in threshold (T) and comfort (C) levels for electrical stimulation, and preferred microphone sensitivity, volume control, and noise-reduction settings, it is essential for the clinician and recipient to determine what combination is best for the individual over several sessions. The results of this study clearly show the advantage of using higher microphone sensitivity settings than the default MS 8 to provide better speech recognition for low-level stimuli. However, it was also necessary to adjust other parameters such as map C levels, automatic sensitivity control and base level, to optimize loudness comfort in the diversity of listening situations an individual encounters in everyday life.

Evaluation of streamlined programming procedures for the Nucleus cochlear implant with the Contour electrode array.

Objective: The objective of this study was to evaluate streamlined programming procedures for the Nucleus cochlear implant system with the Contour electrode array. Design: Phase 1 involved an examination of the clinical MAPs for the first 103 recipients implanted with the Contour electrode array in the Melbourne Cochlear Implant Clinic, to examine the ability to predict the entire MAP based on a smaller number of clinically determined T- and/or C-levels. In phase 2, a subset of the streamlined procedures was selected and clinically evaluated, using speech perception and subjective preference measures. In the first study, the clinical MAP was compared with a MAP based on interpolating across three behavioral T-levels and three behavioral C-levels in a group of newly implanted subjects. The second study investigated the use of a single interpolated profile as the basis to creating the entire MAP. Initial evaluation compared the clinical MAP with two streamlined MAPs, one in which the C-level profile was derived from interpolation across a subset of T-levels and one in which the T-level profile was derived from interpolation across a subset of C-levels. In this case, the interpolated profile was based on five behavioral measures. Subsequently, the use of either three or a single T-level measure as the basis for the interpolated T-level profile was evaluated. Eighteen subjects, who were experienced with the clinical MAP before enrollment in the study, participated in the initial evaluation. The subjects were selected to include a group whose RMS deviation from clinical MAP levels, as determined in Phase 1, was greater than that of the wider population. Results: The Phase 1 analysis showed that as expected, larger differences were observed between the clinical and derived MAP levels as interpolation was applied across fewer measured electrodes and that the use of a single interpolated profile to create the entire MAP resulted in the greatest deviation. No significant group mean difference was found in speech perception scores for newly implanted subjects when mapped with the clinical versus the streamlined MAP based on three behavioral T- and three behavioral C-level measures. For some individual subjects, scores were higher with the streamlined MAP. Subjective reports from the comparative performance questionnaire were consistent with these findings. No significant group mean difference in speech perception scores was found in comparing the clinical MAP with the streamlined MAPs based on a single interpolated T- or C-level profile created from five behavioral measures. Individual effects were observed; however, there was no consistent finding across subjects. The use of three rather than five behavioral T-level measures in the procedure did not result in significantly lower group mean scores; however, significantly poorer scores were obtained for three of the 10 individual subjects. The use of a MAP based on a single behavioral measure did result in poorer speech perception scores when compared with the MAP based on five behavioral T-level measures. These findings were consistent with subjective results from the performance questionnaires administered to determine preference for program across a range of listening situations. Conclusions: Two streamlined programming procedures are recommended for use in the clinical setting: (1) interpolating across three measured T-levels and three measured C-levels and (2) interpolating across five measured T- or C-levels and using the interpolated profile for fitting of the alternative profile.

Clinical evaluation of higher stimulation rates in the nucleus research platform 8 system.

Objective: The effect on speech perception of using higher stimulation rates than the 14.4 kHz available in the Nucleus 24 cochlear implant system was investigated. The study used the Nucleus Research Platform 8 (RP8) system, comprising the CI24RE receiver-stimulator with the Contour electrode array, the L34SP body-worn research speech processor, and the Nucleus Programming Environment (NPE) fitting and Neural Response Telemetry (NRT) software. This system enabled clinical investigation of higher stimulation rates before an implementation in the Freedom cochlear implant system commercially released by Cochlear Limited. Design: Use of higher stimulation rates in the ACE coding strategy was assessed in 15 adult subjects. An ABAB experimental design was used to control for order effects. Program A used a total stimulation rate of between 12 kHz and 14.4 kHz. This program was used for at least the first 3 mo after initial device activation. After evaluation with this program, each subject was provided with two different higher stimulation rate programs: one with a total stimulation rate of 24 kHz and the other with a total stimulation rate of 32 kHz. After a 6-week period of familiarization, each subject identified his/her preferred higher rate program (program B), and this was used for the evaluation. Subjects then repeated their use of program A for 3 wk, then program B for 3 wk, before the second evaluation with each. Speech perception was evaluated by using CNC open-set monosyllabic words presented in quiet and CUNY open-set sentences presented in noise. Preference for stimulation rate program was assessed via a subjective questionnaire. Threshold (T)- and Comfortable (C)-levels, as well as subjective reports of tinnitus, were monitored for each subject throughout the study to determine whether there were any changes that might be associated with the use of higher stimulation rates. Results: No significant mean differences in speech perception results were found for the group between the two programs for tests in either quiet or noise. Analysis of individual subject data showed that five subjects had significant benefit from use of program B for tests administered in quiet and for tests administered in noise. However, only two of these subjects showed benefit in both test conditions. One subject showed significant benefit from use of program A when tested in quiet, whereas another showed benefit with this program in noise. Each subject’s preferred program varied. Five subjects reported a preference for program A, eight subjects reported a preference for program B and two reported no overall preference. Preference between the different stimulation rates provided within program B also varied, with 10 subjects preferring 24 kHz and five preferring 32 kHz total stimulation rates. A significant increase in T-levels from baseline measures was observed after three weeks of initial experience with program B, however there was no difference between the baseline levels and those obtained after five weeks of use. No significant change in C-levels was found over the monitoring period. No long-term changes in tinnitus that could be associated with the use of the higher stimulation rates were reported by any of the subjects. Conclusions: The use of higher stimulation rates may provide benefit to some but not all cochlear implant recipients. It is important to optimize the stimulation rate for an individual to ensure maximal benefit. The absence of any changes in T- and C-levels or in tinnitus suggests that higher stimulation rates are safe for clinical use.

Parameter selection and programming recommendations for the ACE and CIS speech-processing strategies in the Nucleus 24 cochlear implant system

Abstract This study evaluated a range of programming parameters available for the ACE and CIS speech-processing strategies in the Nucleus 24 cochlear implant system. Specifically, the effect on speech perception of adjustments to the number of channels and the stimulation rate in the CIS strategy, and the effect of adjustments to the number of maxima in the ACE strategy were investigated in a group of adult subjects. Based on these findings, and the results of a previous study (Vandali et al., 2000), a number of recommendations for programming were identified that could improve efficiency in the clinical setting where time may be limited. The results suggest that speech perception benefits may be maximized for an individual cochlear implant recipient through concentration on selection of the appropriate stimulation rate. When using the CIS strategy, the number of channels should also be optimized. Adjustment to the number of maxima in the ACE strategy was found to be less likely to provide improvements in speech recognition for a given individual when the number of maxima parameter is set to eight or 12.

Speech perception studies using a multichannel electrotactile speech processor, residual hearing, and lipreading

Three studies are reported on the speech perception of normally hearing and hearing-impaired adults using combinations of visual, auditory, and tactile input. In study 1, mean scores for four normally hearing subjects showed that addition of tactile information, provided through the multichannel electrotactile speech processor, to either audition alone (300-Hz low-pass-filtered speech) or lipreading plus audition resulted in significant improvements in phoneme and word discrimination scores. Information transmission analyses demonstrated the effectiveness of the tactile aid in providing cues to duration, F1 and F2 features for vowels, and manner of articulation features for consonants, especially features requiring detection and discrimination of high-frequency information. In study 2, six different cutoff frequencies were used for a low-pass-filtered auditory signal. Mean scores for vowel and consonant identification were significantly higher with the addition of tactile input to audition alone at each cutoff frequency up to 1500 Hz. The mean speechtracking rate was also significantly increased by the additional tactile input up to 1500 Hz. Study 3 examined speech discrimination of three hearing-impaired adults. Additional information available through the tactile aid was shown to improve speech discrimination scores; however, the degree of increase was inversely related to the level of residual hearing. Results indicate that the electrotactile aid may be useful for patients with little residual hearing and for the severely to profoundly hearing impaired, who could benefit from the high-frequency information presented through the tactile modality, but unavailable through hearing aids.