Research paperElectric-acoustic pitch comparisons in single-sided-deaf cochlear implant users: Frequency-place functions and rate pitch
Introduction
Pitch is one of the most widely studied and passionately debated perceptual attributes in both acoustically and electrically evoked hearing. Theories on pitch perception have evolved from the earliest place (von Helmholtz, 1863) and temporal models (Wundt, 1880) to more refined spectral (Goldstein, 1973, Terhardt, 1979) and temporal autocorrelation (Licklider, 1959, Meddis and O'Mard, 1997) models, some combining place and temporal aspects (Wever, 1940, Wever and Bray, 1930), and to models based on spatial gradients of neural responses that are phase-locked to the traveling-wave induced basilar membrane motion (Loeb, 2005).
In both acoustically and electrically evoked hearing, it has been shown that temporal rate and spectral place are two orthogonal dimensions of perception, with the first dimension being correlated to pitch, and the second dimension to what is commonly described as timbre (Plomp and Steeneken, 1971, Tong et al., 1983). While for acoustic stimuli temporal and place information covary, in electrically evoked hearing in cochlear implants (CIs) the two dimensions may be manipulated independently from each other. However, in many studies, also involving CI users, the two perceptual dimensions of pitch and timbre are not differentiated, primarily because subjects rarely can make a clear distinction between the two attributes. Thus, for simplicity, in the following the term ‘pitch’ will be used to describe sensations that allows listeners to order stimuli from ‘low’ to ‘high’, bearing in mind that both perceptual dimensions of pitch and timbre are likely to contribute to those sensations.
In normal-hearing listeners, Oxenham et al. (2004) investigated whether pitch perception is consistent with a purely temporal model or whether the place code is also an important component in the neural representation of periodic sounds. Pitch perception was compared between normal and ‘transposed’ acoustic stimuli, in which low-frequency temporal information was presented to high-frequency regions of the cochlea. This was accomplished by modulating a low-frequency half-wave rectified sinusoid on top of a high-frequency sinusoidal carrier. For all transposed stimuli, pitch perception was heavily compromised: frequency difference limens were significantly larger for transposed than for normal pure tones, F0 difference limens at 100 Hz for complex tones composed of harmonics 3–5 were unmeasurable in three of four subjects when the harmonics had been transposed, and the same complex tones could not be matched to pure tones at the fundamental frequency when they had transposed harmonics. The authors concluded that the tonotopic representation, i.e. the cochlear place where temporal information is presented, is crucial to complex pitch perception, suggesting that for periodic sounds temporal information ought to be presented at the right tonotopic place in order to elicit a salient pitch percept.
As more patients with residual hearing or unilateral hearing loss benefit from a CI, there are growing opportunities to compare pitch percepts elicited by electrical stimulation through the implant to pitch percepts elicited by acoustic stimuli in the same or contralateral ear. Such comparisons are relevant from both a practical point of view in terms of CI sound processor mapping and from a more fundamental point of view in terms of sound coding strategy design. In practical terms, an allocation of frequency bands in a multi-channel CI sound processor to electrodes in a way that the allocated spectral information closely matches the tonotopic electrode place might not only lead to a better acceptance of the CI sound, but also allow implant patients to reach asymptotic levels of speech perception faster after first implantation (Reiss et al., 2008). Recent studies on sound coding strategies have investigated possibilities to transmit temporal information in general, and low-frequency temporal fine structure in particular, more effectively through a CI (Müller et al., 2012, Riss et al., 2008, Schatzer et al., 2010). Temporal fine structure information may be crucial for the perception of speech in complex backgrounds (Qin and Oxenham, 2003) and tonal languages (Xu and Pfingst, 2003), as well as for the localization of sounds and perception of pitch (Smith et al., 2002). However, the results by Oxenham et al. (2004) suggest that the hypothesized benefits of representing additional temporal fine structure information might be limited unless that information is presented at the correct tonotopic place along the cochlea.
Several studies have investigated pitch percepts across acoustic and electric stimulation modalities in users of different cochlear implant systems (Baumann and Nobbe, 2006, Baumann et al., 2011, Blamey et al., 1996, Boëx et al., 2006, Carlyon et al., 2010b, Dorman et al., 2007, McDermott et al., 2009, Vermeire et al., 2008). Many of these studies found that the pitch elicited through stimulation of intracochlear electrodes was generally between one and two octaves below Greenwood's estimate (1961, 1990) for the frequency-place function in humans (Blamey et al., 1996, Boëx et al., 2006, Dorman et al., 2007). Blamey et al. (1996) conducted pitch-comparison experiments in 13 subjects with relatively poor hearing in their non-implanted ear. Results were quite variable across subjects, and the pitch elicited through stimulation of intracochlear electrodes was generally between Greenwood's prediction and three octaves below that prediction. Boëx et al., 2006, Baumann and Nobbe, 2006, and Dorman et al. (2007) tested subjects that had better hearing thresholds in the non-implanted ear. Thus, pitch matching data were less compromised by hearing loss and abnormal cochlear function. When frequency-place maps were constructed, matches were in a range between Greenwood's prediction and two octaves below.
With the exception of one unilaterally deaf CI subject in Baumann et al. (2011), place-pitch matches in patients with normal or near-normal hearing in the non-implanted ear did not deviate consistently from Greenwood's prediction (Carlyon et al., 2010b, Vermeire et al., 2008). Vermeire et al. performed cross-modality pitch scaling experiments in 14 subjects with functional hearing in the non-implanted ear. They found that electrical stimulation produced a frequency-place function that, on average, resembles Greenwood's function, although results were also showing a large variability across subjects. In the study by Carlyon et al., four CI users with normal hearing in the non-implanted ear compared pitch percepts of electrical and acoustic stimuli presented to the two ears. Results of these comparisons did not show a deviation of electrical pitch percepts from the predictions of Greenwood's cochlear frequency-place equation. Another important observation in that report is that stimulus comparisons across electric and acoustic modalities are adversely affected by differences in perceptual quality, becoming highly susceptible to non-sensory biases. As a consequence, substantial range effects were encountered for all of the applied cross-modality comparison procedures. By carefully examining results for such range biases and discarding pitch matches that did not pass strict ‘sanity’ checks, Carlyon et al. found very little variability across subjects. While Carlyon et al. derived electrical place-pitch matches for electrode positions up to 360° from the round window, to our knowledge only the study by Vermeire et al. has obtained second-turn electrode matches from a larger number of subjects with near-normal contralateral hearing (however, without applying checks for non-sensory biases as in the experiment presented here).
In the present study, we conducted electric-acoustic pitch matching experiments in eight experienced MED-EL implant users having near-normal hearing in the non-implanted ear. All subjects were part of the larger group participating in the study by Vermeire et al. (2008). In the first experiment, frequency-place functions were determined for high-rate unmodulated trains of biphasic pulses presented in monopolar configuration on individual electrodes, including second-turn electrodes. In contrast to the pitch scaling procedure that was used in the 2008 study, in the present study we used an adaptive matching procedure and applied ‘sanity’ checks similar to those proposed by Carlyon et al. (2010b) in order to identify reliable pitch matches. As a result, we expected to find less variability in the frequency-place functions across subjects. Of interest was also the question whether frequency-place functions in our long-electrode subjects would show a systematic shift from the prior study, similar to other observations in short-electrode hybrid subjects (Reiss et al., 2007).
As previously mentioned, recent attempts in implant coding strategies have been made to represent temporal fine structure information in the electrical stimulation patterns delivered to the auditory nerve, in addition to the temporal information already conveyed with envelope stimulation as in the Continuous Interleaved Sampling (CIS) strategy (Wilson et al., 1991). Low-frequency temporal fine structure for instance may be represented via pulse rate modulations on all or only on select apical channels. The effect of electrode place on the perceived pitch produced by low-rate pulse trains was investigated in a second experiment. Based on the findings by Oxenham et al. (2004), it is hypothesized that similar to acoustically evoked hearing, also in electrically evoked hearing tonotopic place and temporal rate of stimulation have to match in order to elicit reliable low-frequency pitch percepts. Seven of the eight subjects participating in the first experiment also compared low-frequency pitch percepts between acoustic and electrical stimulation. The pitch of unmodulated pulse trains presented to one of the six electrodes in the apical half of the array was matched, by adjustments of pulse rates, to that of loudness-balanced pure tones presented to the normal ear at fixed frequencies ranging from 100 to 450 Hz. The goal was to identify combinations of electrode insertion angles and pulse rates that would produce reliable pitch matches to the pure-tones references.
Section snippets
Subjects
Eight subjects who were part of the group reported on by Vermeire et al. (2008) participated in the present study. All are adults and suffered from unilateral severe tinnitus resulting from ipsilateral sensorineural deafness of varying etiologies (Table 1). Participating in a previous study on the effectiveness of cochlear implantation in treating unilateral tinnitus (Van de Heyning et al., 2008), they were all benefitting from a significant reduction of their tinnitus level through stimulation
Subjects
All subjects but S3 from the group participating in experiment 1 also completed this experiment.
Stimuli
As in experiment 1, the acoustic stimuli presented to the normal ear were pure tones with 500 ms duration and linear on- and offset ramps of 25 ms each. Pure tone frequencies were fixed and included 100, 150, 200, 300, and 450 Hz. 450 Hz was chosen as the highest frequency to span the typical range of temporal pitch in electrical hearing, where pitch as a function of rate tends to saturate at
Conclusions
Eight MED-EL CI users, implanted with a 31-mm FLEXSOFT or 24-mm medium M electrode array and with near-normal hearing in their non-implanted ear, compared pitch percepts produced by unmodulated electrical pulse trains and acoustic pure tones. Results were obtained via adaptive matching procedure, with individual tracks starting at frequencies both higher and lower in pitch, and were checked for reliability by discarding non-converging tracks.
Frequency-place functions were derived from tone
Acknowledgments
We thank Dr. David Landsberger for his helpful comments on a previous version of this manuscript and Dr. Otto Peter for providing the RIB II MED-EL implant research interface. Special thanks go to our subjects for their time and commitment. This work was funded by the Austrian C. Doppler Research Association and supported by the TOPBOF project of the University of Antwerp.
References (41)
- et al.
The cochlear implant electrode-pitch function
Hear. Res.
(2006) - et al.
Pitch comparisons of acoustically and electrically evoked auditory sensations
Hear. Res.
(1996) Calculating virtual pitch
Hear. Res.
(1979)- et al.
Neural tonotopy in cochlear implants: an evaluation in unilateral cochlear implant patients with unilateral deafness and tinnitus
Hear. Res.
(2008) - et al.
Pitch matching psychometrics in electric acoustic stimulation
Ear. Hear.
(2011) - et al.
Acoustic to electric pitch comparisons in cochlear implant subjects with residual hearing
J. Assoc. Res. Otolaryngol.
(2006) - et al.
Effect of stimulus level and place of stimulation on temporal pitch perception by cochlear implant users
J. Acoust. Soc. Am.
(2010) - et al.
Pitch comparisons between electrical stimulation of a cochlear implant and acoustic stimuli presented to a normal-hearing contralateral ear
J. Assoc. Res. Otolaryngol.
(2010) - et al.
The pitch of electrically presented sinusoids
J. Acoust. Soc. Am.
(1994) - et al.
An electric frequency-to-place map for a cochlear implant patient with hearing in the nonimplanted ear
J. Assoc. Res. Otolaryngol.
(2007)
Relative importance of rate and place: experiments using pitch scaling techniques with cochlear implants recipients
Ann. Otol. Rhinol. Laryngol. Suppl.
The ideal extent of apical stimulation: where to stop?
An optimum processor theory for the central formation of the pitch of complex tones
J. Acoust. Soc. Am.
A cochlear frequency-position function for several species–29 years later
J. Acoust. Soc. Am.
Critical bandwidth and the frequency coordinates of the basilar membrane
J. Acoust. Soc. Am.
Temporal pitch perception at high rates in cochlear implants
J. Acoust. Soc. Am.
Transformed up-down methods in psychoacoustics
J. Acoust. Soc. Am.
Three auditory theories
Are cochlear implant patients suffering from perceptual dissonance?
Ear. Hear.
Electro-acoustic stimulation. Acoustic and electric pitch comparisons
Audiol. Neurootol.
Cited by (0)
- 1
Present address: Thomas More University College, Jozef De Bomstraat 11, 2018 Antwerpen, Belgium.
- 2
Present address: Klinikum der Universität München, Marchioninistraβe 15, 81377 München, Germany.
- 3
Present address: MED-EL GmbH, Fürstenweg 77a, 6020 Innsbruck, Austria.