Fundamentals of Clinical ECAP Measures in Cochlear Implants--Part 1: Use of the ECAP in Speech Processor Programming

Introduction

Telemetry capabilities have been commercially available since 1998 for measuring the electrically evoked compound action potential (ECAP) from the auditory nerve in cochlear implant recipients. An initial body of research provided foundational information about how to use ECAP measures to assist with speech processor programming for one specific device (Nucleus 24M) and one specific speech processing strategy (SPEAK). Research conducted since that time has focused on refining the clinical use of ECAP threshold measures for speech processor programming with other devices, strategies, and stimulation rates. Now that ECAP measures are readily available in the clinical software for two of the three cochlear implant manufacturers, more clinicians are using these measures as part of routine clinical service delivery. The purpose of this article is to provide an overview of how telemetry is used to measure the ECAP, provide a summary of the two methods used to separate neural response from artifact, and summarize results from a number of past and present research studies that demonstrate how ECAP measures can be used clinically to guide speech processor programming for various devices and processing strategies.

Overview of Telemetry and the ECAP

All newer cochlear implant systems are equipped with two-way telemetry capabilities that allow for quick and easy measurement of electrode impedance and the ECAP. Telemetry simply means data transmission via radio frequency from a source to a receiving station. Neural Response Telemetry (NRT) is the ECAP telemetry software used with the newer Cochlear Corporation devices (Nucleus 24M, 24R(CS) "Contour", 24R(ST), and 24RE "Freedom"). NRT was first introduced in 1996 and FDA approved in 1998 as its own software application separate from the clinical programming software. Today, NRT has been incorporated into the Nucleus Custom Sound clinical programming software so that ECAP threshold information can easily be used to aid in creating speech processor programs, or MAPs. Neural Response Imaging (NRI) is the ECAP telemetry feature for the newer Advanced Bionics devices (Clarion CII or HiRes 90K). It was first introduced in 2001 and FDA approved in 2003 as an integrated part of the SoundWave clinical programming software. Auditory Response Telemetry, or ART, is the ECAP telemetry feature for the newest Med-El device (Pulsar CI100), which has not yet received FDA approval and thus is not currently commercially available.

The ECAP represents a synchronous response from electrically stimulated auditory nerve fibers, and is essentially the electrical version of Wave I of the ABR. The ECAP is recorded as a negative peak at about 0.2-0.4 ms following stimulus onset, followed by a much smaller positive peak or plateau occurring at about 0.6-0.8 ms (Abbas et al., 1999; Brown, Abbas & Gantz, 1998; Cullington, 2000). The amplitude of the ECAP can be as large as 1-2 mV, which is roughly an order of magnitude larger than the electrically evoked auditory brainstem response (EABR) (Brown et al., 1998). ECAP measures have become a popular alternative to clinical EABR testing due to ease of recording and reduced testing time. In contrast to the EABR, ECAP measures do not require surface recording electrodes, sleep/sedation, or additional averaging equipment. The ECAP is recorded via the intracochlear electrodes of the implant, so the neural potential is larger than the EABR and thus fewer averages are needed, which significantly reduces testing time. Because the ECAP is an early-latency evoked potential, there are challenges associated with separating it from stimulus artifact. NRT and NRI use different methods to separate the neural response from stimulus artifact. NRT uses a subtraction method, which exploits refractory properties of the auditory nerve. NRI uses an alternating stimulus polarity method, similar to that used clinically for auditory brainstem response (ABR) testing. It is helpful to understand how each of these artifact reduction methods works in order to make informed decisions about optimizing stimulation and recording parameters and interpreting results.

The Subtraction Method (NRT)

The subtraction method is described in detail in several publications (e.g., Abbas et al., 1999; Brown, Abbas & Gantz, 1990, 1998; Brown et al., 2000; Dillier et al., 2002). With the most recent implementation of the subtraction method, four stimulus conditions are presented: (1) probe stimulus alone, (2) masker stimulus followed by probe, (3) masker stimulus alone, and (4) zero-current pulse to measure system artifact. However, for simplicity sakes, the first two conditions are the most important for understanding how this method works. In the probe-alone condition (condition 1), the probe recruits auditory nerve fibers. The measured waveform contains a neural response as well as stimulus artifact. In the masker-plus-probe condition (condition 2), the masker stimulus recruits the same auditory nerve fibers as in condition 1 (if the stimuli are presented on the same electrode at the same level). The probe is then presented shortly after the masker (typically 0.5 ms later), within the refractory period of those fibers. As a result, only stimulus artifact (no neural response) is recorded for the probe stimulus in condition 2. If the response to the probe in condition 2 (i.e., artifact only) is subtracted from the response to the probe in condition 1 (i.e., neural response and artifact), then the neural response to the probe can be isolated from the stimulus artifact. The last two stimulus conditions in the 4-step process (masker alone and system artifact) are collected and subtractions applied to remove additional artifacts and neural response to the masker in condition 2.

Because the subtraction method in NRT employs a 4-step stimulus paradigm, there are a few stimulating and recording parameters that can be manipulated for optimizing recordings, which are not available or necessary with the alternating polarity method used in NRI. The first parameter is the recording delay. This is a user-controlled time period between offset of the probe pulse and the start of recording the response (Dillier et al., 2002). If the recording delay is too short, then too much artifact will show up in the recorded response. If the recording delay is too long, then the negative peak will not be resolved, making it impossible to measure a peak-to-peak ECAP amplitude. The second parameter unique to the subtraction method is the masker advance (MA) or inter-pulse interval (IPI). This is the time between offset of the masker and onset of the probe in the masker-plus-probe condition. The default value for the IPI is 0.5 ms, as this has been shown to be the optimal time delay where most auditory nerve fibers are fully within their refractory period (e.g., Brown & Abbas, 1990; Brown, Abbas & Gantz. 1990, 1998; Miller, Abbas & Brown, 2000). For routine clinical ECAP measures, the IPI is generally not changed unless the user is specifically measuring IPI effects (e.g., to choose a fast versus slow stimulation rate for a given processing strategy) (Shpak, Berlin & Luntz, 2004). Part 2 of the present article, coming in November 2006 on Audiology Online, will further discuss stimulation and recording parameter changes as well as offer guidance on how to pick peaks and determine ECAP threshold.

The Alternating Polarity Method (NRI)

The alternating polarity method is used routinely in clinical ABR testing. With this method, responses are measured for negative-leading (cathodic) and positive-leading (anodic) biphasic pulses. As we know with acoustically evoked potentials, switching the polarity of the stimulus results in switching polarity of the artifact. However, the polarity of the neural response remains the same. Upon averaging responses from both polarities, the majority of the stimulus artifact cancels out, leaving the neural response. There is one hitch to this method though: ECAP amplitude and latency are different for cathodic versus anodic electrical stimulation (Miller et al., 1998). Thus, the averaged response amplitude and morphology may be very different than that of either polarity alone. As a result, the averaged waveform for the alternating polarity method tends to be significantly smaller than that obtained with the subtraction method (r = 0.97, p
Common Clinical Uses for ECAP Measures

Over the past five or six years there has been a steady increase in the number of studies addressing the clinical utility of ECAP measures. In general, ECAP measures are clinically useful for a number of applications, including:

• Objective verification of auditory nerve function in response to electrical stimulation
  
  
• Objective verification of device function
  
  
• Assistance in programming the speech processor for individuals who cannot provide reliable behavioral responses
  
  
• Verification or confirmation of the accuracy of questionable behavioral responses

The focus of the majority of research in this area has been on investigating how ECAP measures can be used clinically to guide speech processor programming. NRT was the first commercially available software to measure the ECAP, and was first implemented with the Nucleus 24M device. Thus, the earliest works compared ECAP thresholds obtained with NRT to behavioral thresholds (T-levels) and comfort levels (C-levels) for the 250-pps SPEAK strategy (default 25 ï ­sec/phase pulse width) in Nucleus 24M straight array cochlear implant recipients (Brown et al., 1998, 2000; Cullington, 2000; Franck & Norton, 2001; Franck, Shah, Hayman, Peterson & Marsh, 2001; Hughes, Brown, Abbas, Wolaver & Gervais, 2000; Hughes, Vander Werff, et al., 2001; Shallop, Facer & Peterson, 1999; Smoorenburg, Willeboer & van Dijk, 2002; Thai-Van et al., 2001). Those studies found the following important outcomes:

Nucleus 24M, 250-pps SPEAK:

• For children, ECAP thresholds generally fall between T- and C-level, thus representing a level that should be audible but not uncomfortable (Di Nardo, Ippolito, Quaranta, Cadoni & Galli, 2003; Hughes et al., 2000).
  
  
• For adults, ECAP thresholds also generally fall between T- and C-level; however, for roughly 1/3 of this population ECAP thresholds may exceed C-level (Brown et al., 1998, 2000; Cullington, 2000; Smoorenburg et al., 2002; Franck & Norton, 2001).
  
  
• T-levels typically fall 22-29 manufacturer-defined current level (CL) programming units below ECAP threshold for children and adults (Cullington, 2000; Hughes et al., 2000; Polak, Hodges & Balkany, 2005). C-levels typically fall 4-10 CL above ECAP threshold for adults (Cullington, 2000; Hughes et al., 2000; Polak et al., 2005) and 21 CL above ECAP threshold for children (Hughes et al., 2000).
  
  
• ECAP thresholds typically fall within the same percentage of MAP dynamic range across electrodes within a subject (Hughes et al., 2000). For children, ECAP thresholds fell on average at 53% of the MAP dynamic range (Di Nardo et al., 2003; Hughes et al., 2000). For adults, ECAP thresholds fell closer to C-level: average at 91% of the MAP dynamic range (Brown et al., 2000; Di Nardo et al., 2003; Hughes et al., 2000).
  
  
• In many cases, ECAP thresholds follow a similar contour or shape to the MAP. When the contour of MAP levels across electrodes is different for T-levels versus C-levels within a subject, the ECAP will often mirror the shape of one of those functions, usually T-level (Hughes, Vander Werff et al., 2001).
  
  
• Smoorenburg et al. (2002) showed that the tilt of the ECAP contour was significantly correlated with the tilt of the T-level contour, but not with the C-level contour. This finding suggests that ECAP measures are better suited for predicting T-levels than C-levels. This conclusion was also supported by other studies (e.g., Franck & Norton, 2001).
  
  
• There is a moderate correlation between ECAP threshold and T-levels. Correlation coefficients vary across studies, ranging between r = 0.5 to 0.9 (Brown et al., 2000; Cullington, 2000; Di Nardo et al., 2003; Franck & Norton, 2001; Hughes et al., 2000; Polak et al., 2005; Smoorenburg et al., 2002; Thai-Van et al., 2001). The correlation between ECAP thresholds and C-levels varies more widely across studies, ranging from r = 0.1 to 0.9 (Brown et al., 2000; Cullington, 2000; Di Nardo et al., 2003; Franck & Norton, 2001; Hughes et al., 2000; Polak, et al., 2005; Smoorenburg et al., 2002; Thai-Van et al., 2001). These results suggest that ECAP measures alone are not reliable enough to set MAP levels directly. However, when ECAP data on all electrodes are combined with a limited amount of behavioral information, there is a significant correlation (r = 0.8 to 0.9) between measured and predicted T- and C-levels (Brown et al., 2000; Franck & Norton, 2001; Hughes et al., 2000).

Since that initial body of research was published, electrode designs have changed to emphasize perimodiolar position (e.g., addition of electrode positioning systems or pre-curved electrode arrays), the stimulation rate of speech processing strategies has increased, and NRI (using alternating polarity) was introduced for use with the Clarion system. Because all of these variables have an effect on behavioral threshold and/or ECAP measures, additional research has been conducted to further define the relation between ECAP thresholds and behavioral levels for these other variables (e.g., Eisen & Franck, 2004; Han et al., 2005; Polak et al., 2005). These further studies are important because perimodiolar electrode position has been shown to result in lower T- and/or C-levels when compared to traditional straight electrode arrays (Kreft, Donaldson, & Nelson, 2004; Parkinson et al., 2002; Saunders et al., 2002; Young & Grohne, 2001), which may (or may not) change the relation between behavioral levels and ECAP measures. In addition, behavioral threshold and comfort levels tend to decrease with increased stimulation rate due to effects of temporal integration (e.g., Brown, Hughes, Lopez & Abbas, 1999; Skinner, Holden, Holden & Demorest, 2000). This will likely also affect the relation between behavioral levels and ECAP measures, because ECAP measures cannot be made using the faster stimulation rates that are used for behavioral levels. When the stimulation rate is increased, ECAP amplitude decreases (and thus threshold increases) due to neural adaptation. Thus we might expect that ECAP thresholds (obtained with traditional slow-rate stimuli) would be more likely to approximate or exceed C-level for these faster-rate strategies. Lastly, ECAP thresholds have been shown to differ when measured using alternating polarity versus a forward masking subtraction method (Eisen & Franck, 2004; Hughes et al., 2003). Although research comparing ECAP thresholds to behavioral levels for different electrode designs, faster rates, and ECAP measurement techniques (e.g., NRI) is still somewhat limited, results from more recent studies found the following notable outcomes for the use of ECAP measures to assist with programming of newer cochlear implant technology:

Perimodiolar electrode positioning:

• In general, ECAP thresholds, T-level and/or most comfortable levels (M-levels) are lower for Clarion electrode arrays with a positioner than without (Donaldson et al., 2001; Eisen & Franck, 2004; Franck et al., 2001; Kreft et al., 2004; Young & Grohne, 2001).
  
  
• Results vary across studies as to whether there are significant differences in T-level, C-level and/or ECAP threshold between the Nucleus 24M straight array and the perimodiolar Nucleus 24R(CS) Contour array. Some studies found significantly lower T- and C-levels in the Contour than in the straight array (Parkinson et al., 2002; Saunders et al., 2002), whereas other studies found no difference in T-level, C-level, or ECAP threshold between the two device types (Hughes, Abbas, Brown, Seyle & South, 2001; Polak et al., 2005).
  
  
• There is a significant correlation between ECAP threshold and both T-levels (r = 0.69) and C-levels (r = 0.76) for a 250 pps rate in the perimodiolar Nucleus Contour array (Polak et al., 2005). This result is similar to earlier findings reported for the Nucleus 24M straight array.

Faster stimulation rates / NRI:

• ECAP thresholds with NRI tend to fall between T-level and M-level (most comfortable level), but most commonly approximate or even exceed M-level for very fast stimulation rates (i.e., HiResolution processing strategy) (Eisen & Franck, 2004; Han et al., 2005).
  
  
• On average, ECAP thresholds with NRI fell at 76% of M-level for pediatric Hi-Resolution users (Eisen & Franck, 2004) and at 110% of M-level for a group of pediatric and adult Hi-Resolution users (Han et al., 2005).
  
  
• ECAP thresholds with NRT tend to approximate C-level for faster-rate processing strategies (i.e., 900 pps; McKay, Fewster & Dawson, 2005). McKay et al. (2005) have published a helpful plot of equal-loudness curves that can easily be used for a first approximation of T- and C-levels for different rates of stimulation (see Figure 3, McKay et al., 2005).
  
  
• In general, as stimulation rate increases for behavioral levels, the correlation between ECAP threshold and behavioral levels decreases (Murray, 2001 as cited in McKay et al., 2005; Zimmerling & Hochmair, 2002).
  
  
• There is a moderate correlation between ECAP threshold and the fast-rate Hi-Resolution M-levels. Correlation coefficients vary across studies, ranging from r = 0.66 to 0.74 (Eisen & Franck, 2004; Han et al., 2005). Like the results reported for 250-pps SPEAK, when ECAP thresholds are combined with a limited amount of behavioral information, there is a significant correlation between measured and predicted M-levels (r = 0.98; Eisen & Franck, 2004).

The results from these later studies suggest that ECAP thresholds are still useful for speech processor programming for faster-rate strategies, various electrode array designs, and different types of artifact reduction methods or software (i.e., NRT vs. NRI). In general, results are similar to those reported in earlier studies for ECAP measures made in the Nucleus 24M straight array for the slow-rate SPEAK processing strategy. However, as the stimulation rate of the speech processing strategy increases, ECAP thresholds are more likely to approximate the upper comfort level. Additional research is needed to further quantify the relation between ECAP thresholds and behavioral levels for different stimulation rates and strategy types.

Summary

ECAP measures can provide valuable objective information to assist with programming cochlear implant speech processors, regardless of device type or processing strategy used. ECAP measures are quickly and easily made in a clinical setting and can be especially useful when coupled with other objective measures such as the electrically evoked stapedial reflex threshold (ESRT) (e.g., Gordon, Papsin & Harrison 2004a, 2004b; Polak et al., 2005). While the correlations between ECAP thresholds and behavioral levels are generally not strong enough to recommend using ECAP measures as a sole means to program the processor, ECAP thresholds do represent a level at which the stimulus should be audible and probably not uncomfortable. Information from specific studies cited above can provide the clinician with a reasonable starting point for creating a first-approximation MAP and/or to begin conditioning a child to respond behaviorally.

References

Abbas, P. J., Brown, C. J., Shallop, J. K., Firszt, J. B., Hughes, M. L., Hong, S. H., & Staller, S. J. (1999). Summary of results using the Nucleus CI24M implant to record the electrically evoked compound action potential. Ear and Hearing, 20, 45-59.

Brown, C. J. & Abbas, P. J. (1990). Electrically evoked whole-nerve action potentials. II. Parametric data from the cat. Journal of the Acoustical Society of America, 88, 2205-2210.

Brown, C. J., Abbas, P. J., & Gantz, B. (1990). Electrically evoked whole-nerve action potentials: Data from human cochlear implant users. Journal of the Acoustical Society of America, 88 (3), 1385-1391.

Brown, C. J., Abbas, P. J., & Gantz, B. (1998). Preliminary experience with Neural Response Telemetry in the Nucleus CI24M cochlear implant. The American Journal of Otology, 19, 320-327.

Brown, C. J., Hughes, M. L., Lopez, S. M., & Abbas, P. J. (1999). Relationship between EABR thresholds and levels used to program the Clarion speech processor. Annals of Otology, Rhinology & Laryngology, 108 (Suppl. 177), 50-57.

Brown, C. J., Hughes, M. L., Luk, B., Abbas, P. J., Wolaver, A., & Gervais, J. (2000). The relationship between EAP and EABR thresholds and levels used to program the Nucleus 24 speech processor: Data from adults. Ear and Hearing, 21, 151-163.

Cullington, H. (2000). Preliminary neural response telemetry results. British Journal of Audiology, 34, 131-140.

Di Nardo, W., Ippolito, S., Quaranta, N., Cadoni, G., & Galli, J. (2003). Correlation between NRT measurement and behavioural levels in patients with the Nucleus 24 cochlear implant. Acta Otorhinolaryngologica Italica, 23, 352-355.

Dillier, N., Lai, W. K., Almqvist, B., Frohne, C., Müller-Deile, J., Stecker, M., & von Wallenberg, E. (2002). Meaurement of the electrically evoked compound action potential via a neural response telemetry system. Annals of Otology, Rhinology & Laryngology, 111 (5), 407-414.

Donaldson, G. S., Peters, M. D., Ellis, M. R., Friedman, B. J., Levine, S. C., & Rimell, F L. (2001). Effects of the Clarion electrode positioning system on auditory thresholds and comfortable loudness levels in pediatric patients with cochlear implants. Archives Otolaryngology Head and Neck Surgery, 127, 956-960.

Eisen & Franck, (2004). Electrically evoked compound action potential amplitude growth functions and HiResolution programming levels in pediatric CII implant subjects. Ear and Hearing, 25, 528-538.

Franck, K. H. & Norton, S. J. (2001). Estimation of psychophyical levels using the electrically evoked compound action potential measured with the neural response telemetry capabilities of Cochlear Corporation's CI24M device. Ear and Hearing, 22, 289-299.

Franck, K., Shah, U., Hayman, C., Peterson, J., & Marsh, R. (2001). Effects of electrode design on mapping levels in children. Poster presented at the 8th Symposium on Cochlear Implants in Children, February 28-March 3, 2001, Los Angeles, CA.

Frijns, J. H. M, Briaire, J. J., de Laat, J. A. P. M., & Grote, J. J. (2002). Initial evaluation of the Clarion CII cochlear implant: Speech perception and neural response imaging. Ear and Hearing, 23, 184-197.

Gordon, K. A., Papsin, B. C., & Harrison, R. V. (2004a). Toward a battery of behavioral and objective measures to achieve optimal cochlear implant stimulation levels in children. Ear and Hearing, 25, 447-463.

Gordon, K. A., Papsin, B. C., & Harrison, R. V. (2004b). Programming cochlear implant stimulation levels in infants and children with a combination of objective measures. International Journal of Audiology, 43, S28-S32.

Han, D.-M., Chen, X.-Q., Zhao, X.-T., Kong, Y. Li, Y.-X., Liu, S., Liu, B., & Mo, L.-Y. (2005). Comparisons between Neural Response Imaging thresholds, electrically evoked auditory reflex thresholds and most comfortable loudness levels in CII Bionic Ear users with HiResolution sound processing strategies. Acta Oto-Laryngologica, 125, 732-735.

Hughes, M. L., Abbas, P. J., Brown, C. J., Etler, C., Behrens, A., & Dunn, S. (2003). "Comparison of two methods used to measure the ECAP in subjects implanted with the Clarion CII device." Paper presented at the Third International Symposium and Workshops on Objective Measures in Cochlear Implants, Ann Arbor, MI, June 26-28, 2003.

Hughes, M. L., Abbas, P. J., Brown, C. J., Seyle, K., & South, H. (2001). Assessing spatial spread in straight and modiolar hugging electrode arrays. Poster presented at the 8th Symposium on Cochlear Implants in Children, February 28-March 3, 2001, Los Angeles, CA.

Hughes, M. L., Brown, C. J., Abbas, P. J., Wolaver, A. A., & Gervais, J. P. (2000). Comparison of EAP thresholds with MAP levels in the Nucleus 24 cochlear implant: Data from children. Ear and Hearing, 21, 164-174.

Hughes, M. L., Vander Werff, K. R., Brown, C. J., Abbas, P. J., Kelsay, D. M. R., Teagle, H. F. B., & Lowder, M. W. (2001). A longitudinal study of electrode impedance, the electrically evoked compound action potential, and behavioral measures in Nucleus 24 cochlear implant users. Ear and Hearing, 22, 471-486.

Kreft, H. A., Donaldson, G. S., & Nelson, D. A. (2004). Effects of pulse rate on threshold and dynamic range in Clarion cochlear implant users. Journal of the Acoustical Society of America, 115, 1885-1888.

McKay, C. M., Fewster, L., & Dawson, P. (2005). A different approach to using Neural Response Telemetry for automated cochlear implant processor programming. Ear and Hearing, 26, 38S-44S.

Miller, C. A., Abbas, P. J., & Brown, C. J. (2000). An improved method of reducing stimulus artifact in the electrically evoked whole-nerve potential. Ear and Hearing, 21, 280-290.

Miller, C. A., Abbas, P. J., Rubinstein, J. T., Robinson, B. K., Matsuoka, A. J., & Woodworth, G. (1998). Electrically evoked compound action potentials of guinea pig and cat: Responses to monopolar, monophasic stimulation. Hearing Research, 119, 142-154.

Parkinson, A. J., Arcaroli, J., Staller, S. J., Arndt, P. L., Cosgriff, A., & Ebinger, K. (2002). The Nucleus 24 Contour cochlear implant system: Adult clinical trial results. Ear and Hearing, 23, 41S-48S.

Polak, M., Hodges, A., & Balkany, T. (2005). ECAP, ESR and subjective levels for two different Nucleus 24 electrode arrays. Otology & Neurotology, 26, 639-645.

Saunders, E., Cohen, L. Aschendorff, A., Shapiro, W., Knight, M., Stecker, M., Richter, B., Waltzman, S., Tykocinski, M., Roland, T., Laszig, R., & Cowan, R. (2002). Threshold, comfortable level, and impedance changes as a function of electrode-modiolar distance. Ear and Hearing, 23, 28S-40S.

Shallop, J. K., Facer, G. W., & Peterson, A. (1999). Neural response telemetry with the CI24M cochlear implant. The Laryngoscope, 109, 1755-1759.

Shpak, T., Berlin, M., & Luntz, M. (2004). Objective measurements of auditory nerve recovery function in Nucleus CI 24 implantees in relation to subjective preference of stimulation rate. Acta Otolarygologica, 124, 582-586.

Skinner, M. W., Holden, L. K., Holden, T. A., & Demorest, M. E. (2000). Effect of stimulation rate on cochlear implant recipients' thresholds and maximum acceptable loudness levels. Journal of the American Academy of Audiology, 11, 203-213.

Smoorenburg, G. F., Willeboer, C., & van Dijk, J. E. (2002). Speech perception in Nucleus CI24M cochlear implant users with processor settings based on electrically evoked compound action potential thresholds. Audiology & Neurotology, 7, 335-347.

Thai-Van, H., Chanal, J.-M., Coudert, C., Veuillet, E., Truy, E., & Collet L. (2001). Relationship between NRT measurements and behavioral levels in children with the Nucleus 24 cochlear implant may change over time: preliminary report. International Journal of Pediatric Otorhinolaryngology, 58, 153-162.

Young, N. M. & Grohne, K. M. (2001). Comparison of pediatric Clarion recipients with and without the electrode positioner. Otology & Neurotology, 22, 195-199.

Zimmerling, M. J. & Hochmair, E. S. (2002). EAP recordings in Ineraid patients - correlations with psychophysical measures and possible implications for patient fitting. Ear and Hearing, 23, 81-91.