Otology & Neurotology 35:1433Y1439 Ó 2014, Otology & Neurotology, Inc.
Sound Localization Abilities of Unilateral Hybrid Cochlear Implant Users With Bilateral Low-Frequency Hearing *†Marc J. W. Lammers, ‡Thomas Lenarz, *†Gijsbert A. van Zanten, *†Wilko Grolman, and §Andreas Buechner *Department of OtorhinolaryngologyYHead and Neck Surgery, ÞBrain Center Rudolf Magnus, University Medical Center Utrecht, The Netherlands; þCluster of Excellence ‘‘Hearing4All,’’ Department of Otolaryngology, and §Cluster of Excellence ‘‘Hearing4All,’’ German Hearing Center, Hannover Medical School, Germany
Objective: To assess the sound localization abilities of subjects unilaterally implanted with a hybrid cochlear implant in different sound localization conditions. Study Design: A prospective, single-subject repeated measures design was performed to assess the sound localization abilities in 5 different listening conditions: combined (hybrid cochlear implant and contralateral acoustic hearing), bimodal, bilateral acoustic, ipsilateral acoustic, and contralateral acoustic. Setting: Tertiary referral center. Patients: Eighteen subjects with bilateral low-frequency residual hearing, implanted with a hybrid cochlear implant, were enrolled in this study. Main Outcome Measures: Postoperative sound localization in 5 different listening conditions. Results: Average group results showed a significant improvement in sound localization for all binaural hearing conditions
(combined, bimodal, and bilateral acoustic) over monaural conditions (ipsilateral acoustic and contralateral acoustic). Subjects performed significantly better in the combined condition compared with the bimodal and bilateral acoustic conditions when sound was presented from the front and the side of their cochlear implant. Conclusion: Best results, for most subjects, were obtained with the routinely used combined fitting. The additional acoustic stimulation of the implanted ear did not significantly improve sound localization. However, a marked improvement has been found by the addition of electrical stimulation, for especially the azimuths ipsilateral and to the front of the cochlear implant, implying the importance of the cochlear implant in sound localization. Key Words: Cochlear implantVElectro-acoustic stimulationVHearing aidVResidual hearingVSound localization. Otol Neurotol 35:1433Y1439, 2014.
Combining electrical speech processing with lowfrequency residual hearing is increasingly becoming a more common strategy for patients with severe-to-profound high-frequency hearing loss who are unable to obtain satisfactory results with modern hearing aids. Preservation of the residual hearing in these subjects is essential for obtaining the best results. Accumulating evidence reveals that this might be possible by using shorter and less traumatic electrodes and that auditory performance is
comparable or even better than with a conventional cochlear implant (CI) (1Y6). Recent results with this hybrid CI have demonstrated a high probability of postoperative residual hearing preservation and improved auditory performance in the electroacoustic stimulation condition (7). This preserved residual hearing might be especially beneficial in complex auditory conditions, such as speech in background noise or sound localization (1,8). When localizing sounds, traditional CI users mainly rely on interaural level differences (ILD) (9,10). The additional low-frequency acoustic stimulation in patients with residual hearing could provide them with interaural time difference (ITD) cues, which might aid them in their sound localization (8). To date, little is known about the sound localization abilities of patients who have undergone hearing preservation surgery with cochlear implantation and who have 2 ears with sensitivity to low-frequency acoustic stimulation.
Address correspondence and reprint requests to Andreas Buechner, Ph.D., Department of Otolaryngology, Medizinische Hochschule Hannover, Carl Neuberg Strasse 1, DEY30625 Hannover, Germany; E-mail: [email protected] Financial disclosure: This research received no specific grants from any funding agency in the public, commercial, or not-for-profit sectors. Competing Interests: Wilko Grolman received unrestrictive research grants from Cochlear Ltd., Med-El GmbH, and Advanced Bionics. No competing interests declared by Marc Lammers, Thomas Lenarz, Gijsbert van Zanten, or Andreas Buechner.
1433
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
1434
M. J. W. LAMMERS ET AL.
One study reported on the sound localization abilities of subjects implanted with a short electrode and demonstrated that these subjects are able to use their residual hearing for sound localization (8). In this study, we evaluated the sound localization abilities of patients with bilateral residual hearing, who have been fitted with a unilateral hybrid cochlear implant.
(open squares), as well as the average preoperative audiograms for the implanted ear (filled circles), are displayed in Figure 1. A postoperative pure-tone average (PTA) was calculated using the threshold levels measured at 125, 250, 500, 750, and 1,000 Hz. If there was no response up to the limit of the audiometer, a nominal value of 130 dB was assigned for the purpose of calculation, indicating complete hearing loss.
MATERIALS AND METHODS
A prospective single-subject repeated measures design was performed in a cohort of 18 hybrid CI users. All subjects were tested in 5 different conditions to evaluate the localization abilities using the hybrid CI and their contralateral ear with or without hearing aids. Conditions were tested in random order and consisted of the following: combined, bimodal, bilateral acoustic only, ipsilateral acoustic only, and contralateral acoustic only. The combined condition referred to the hybrid CI and acoustic stimulation (AS) in the contralateral ear (CIHA + AS). In the bimodal condition only the CI of the hybrid device with contralateral acoustic stimulation was tested (CI + AS). The HA of the hybrid CI was removed and the implanted ear was plugged and covered with earmuffs to minimize acoustic hearing. In the bilateral acoustic condition, the CI was switched off, so that both ears were solely acoustically stimulated (HA + AS). In the ipsilateral acoustic condition, the contralateral ear was plugged and covered with earmuffs, whereas in the implanted ear, only the HA of the hybrid CI was switched on. In the contralateral acoustic condition, the hybrid CI was removed, and the implanted ear was plugged and covered with earmuffs.
Study Design Subjects Eighteen adults, with bilateral residual hearing and unilaterally implanted with the Cochlear Nucleus Hybrid-L device (Cochlear Corp., Australia), were enrolled in this study. Subject demographics of the subjects are shown in Table 1. Most subjects had experience with the sound localization test. Filter bank settings of the hybrid CI were set to either an overlapping map or a nonoverlapping map. In the overlapping map setting, the full frequency spectrum of approximately 200 to 7,000 Hz is transmitted through the cochlear implant, whereas in the nonoverlapping map setting a reduced frequency range, depending on the residual hearing of the individual subject, is transmitted (11). Twelve subjects used the nonoverlapping filter bank settings, and 13 subjects used a contralateral hearing aid, as indicated in Table 1. The contralateral hearing aids were fitted by hearing aid acousticians and were checked prior to the test procedure for proper functioning. The fitting itself was not changed before testing and their daily settings were used. Cochlear Hybrid Prescription was used for the hearing aids of the hybrid CI. Loudness balancing was done during the fitting of the CI, i.e. the loudness of the CI was adapted such that the perceived loudness in live voice was equal between the left and right ear. The group average postoperative pure tone audiogram for the implanted ear (solid triangles) and the contralateral ear TABLE 1. Age at Experience Subject Age implantation with CI no. (yr) (yr) (mo) 1 2
67,9 74,0
67,7 73,1
1 9
3 4 5 6 7 8 9
26,2 65,3 45,7 73,4 65,7 52,5 67,1
23,3 63,1 44,7 72,8 62,3 51,3 65,7
34 25 11 6 40 13 16
10 11 12 13 14 15 16 17
49,9 50,6 20,0 69,4 58,4 68,3 57,6 70,3
49,0 49,8 17,8 67,7 56,0 66,4 56,9 68,9
10 8 25 20 27 20 6 14
18 Mean
25,8 56,0
24,6 54,5
13 17
Etiology Unknown Sudden deafness Preterm birth Ototoxic Trauma Genetic Unknown Ototoxic Sudden deafness Unknown Unknown Unknown Unknown Rubella Otitis Media Genetic Sudden deafness Unknown
Sound Localization Sound localization tests were performed in a sound-treated booth (12 m2) with 12 loudspeakers spaced equidistantly at 30-degree intervals spanning 360 degrees, located approximately 1 m above the floor, and numbered from 1 to 12 in a clockwise
Subject demographics
Duration of HA Duration of HL Duration of HA Contralateral use contralateral in implanted use in implanted ear ear (yr) ear (yr) ear (yr)
Filter bank setting
LF (Hz)
Overlapping Overlapping
188 188
HA HA
11 9
16 21
11 9
HA No HA HA HA No HA HA No HA
19 40a 22 5 20a 13 19a
24 40 45 8 22 20 15
19 40 36 6 15 13 15
Nonoverlapping 813 Nonoverlapping 688 Nonoverlapping 438 Nonoverlapping 1188 Overlapping 188 Nonoverlapping 438 Nonoverlapping 813
No HA HA HA HA HA HA HA HA
5a 19 17 15 24 53 43 4
1 50 20 15 33 68 57 4
19 17 15 24 53 37 4
Nonoverlapping 688 Nonoverlapping 1688 Overlapping 188 Nonoverlapping 438 Overlapping 188 Nonoverlapping 1188 Nonoverlapping 688 Nonoverlapping 563
No HA
-a 20
25 27
21
a
Overlapping
Subject was not using hearing aids (HAs) at the time of testing. CI indicates cochlear implant; HA, hearing aid; HL, hearing loss; LF, lowest frequency transmitted through the cochlear implant.
Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
188
SOUND LOCALIZATION WITH A HYBRID CI
1435
lateral HA, nonoverlapping map users with contralateral HA, overlapping map users without contralateral HA, and nonoverlapping map users without HA) were analyzed with KruskalWallis and Mann-Whitney U tests. If group differences were not statistically significant, pooling of the data was deemed possible. Differences between the 5 different conditions and the 12 sound source localizations were analyzed using repeated measures ANOVAs. Significant main effects ( p G 0.05) were followed-up with Bonferroni post hoc tests, and the Greenhouse-Geisser correction was applied to compensate for violations of the sphericity assumption. Linear regression analyses were performed to evaluate the influence of variables, such as degree of hearing loss and age at visit on sound localization performance.
RESULTS
FIG. 1. Group average pure tone audiograms (n = 18). The subjects’ most recent postoperative audiogram for the implanted ear (solid triangles) and preoperative audiogram for the implanted ear (filled circles) and contralateral ear (open squares) are presented.
sequence (Fig. 2). The subject was seated in the center of the circular array of loudspeakers, facing the number 12 loudspeaker, at a distance of 1.2 meters. The stimulus consisted of one short sentence from the Hochmair-Schulz-Moser (HSM) sentence test and was presented by one of the 12 loudspeakers, in a random order. The sound level was fixed at a level of 65 dB SPL. No sound intensity roving was applied. This stimulus configuration was used to compare our results with previous sound localization tests with the hybrid CI and conventional CI users, conducted in our center (12,13). Every test consisted of four presentations by all 12 loudspeakers. During the test, patients were instructed to keep their heads steadily directed toward loudspeaker 12. Head movements were visually monitored by the investigator. Subjects were asked to verbally identify the number corresponding to the loudspeaker from which the sentence was presented. The response errors in all test conditions were calculated and converted to degrees and root mean square (RMS) errors. Before the start of the test, subjects were given a short sample, by presenting the sentence from loudspeaker 12, 3, 6, and 9 (0, 90, 180, and j90 degrees). During the test, only one repeat was allowed if subjects could not hear the first stimulus. If a subject failed to complete the test, the chance RMS error of 104 degrees was awarded. No feedback was provided. All subjects used their routinely used filter bank setting and were not allowed to change their map settings. Because of time constraints, the hybrid CI and, if applicable, the contralateral hearing aid were not loudness balanced before the sound localization test. Neither were differences in levels in the monaural and binaural conditions corrected for binaural summation effects. The present study was performed according to the principles of the Declaration of Helsinki and received the approval of the ethical committee of the Medical University of Hannover.
Statistical Analyses Statistical analyses were completed using SPSS version 20.0 software. Group differences between the different patient groups included in this study (i.e., overlapping map users with contra-
Figure 3 displays the results of a subject with a moderate localization performance (upper row) and a subject experiencing difficulties in determining the exact location of a sound (lower row). Scores did not significantly differ between hearing aid and nonhearing aid users for all 5 conditions (Mann-Whitney U tests: p 9 0.22). Results of subjects using the overlapping map did not significantly differ from the results of users with the nonoverlapping maps (Mann-Whitney U tests: p 9 0.21). Because the average results of the 4 groups (overlapping with or without HA and nonoverlapping with or without HA) did not differ significantly, the data obtained from all 18 subjects was pooled. Individual and mean RMS errors for all 18 subjects in the 5 different conditions are presented in Table 2 and Figure 4. Repeated measure analyses indicated that significant differences between the different conditions were present (F4,68 = 19.081; p G 0.001). Bonferroni post hoc analyses revealed a significantly better performance in all
FIG. 2. Sound localization test setup. The speakers were spaced equidistantly at 30-degree intervals spanning 360 degrees at 1.2 meters from the subject. Subject was seated in the center, facing loudspeaker 12. Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
1436
M. J. W. LAMMERS ET AL.
FIG. 3. Example of localization results (in degrees) of 2 subjects (Subjects 3 and 6) for the 3 binaural conditions. A subject with moderate performance is presented on top, and a subject with poor performance in the bottom row. Means and standard deviations (T error bars) of all responses for each of the 12 loudspeaker positions are shown. The diagonal line represents perfect performance. Because of the 360degree test setup, response angles can vary from þ180 to j180 degrees around the source angle, that is, responses for a source angle of 150 degrees can vary from j30 to 330 degrees; 0 to 180 degrees refers to the Hybrid-L side.
binaural conditions (combined, bimodal, and bilateral acoustic) as compared with the monaural conditions (ipsilateral acoustic and contralateral acoustic) (p G 0.05). All conditions performed significantly better than the chance TABLE 2. Subject no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Mean SD
RMS error of 104 degrees (p G 0.01). There were no significant differences found between both monaural conditions. A nonsignificant RMS error difference of 9 degrees was found between the combined and the bimodal condition
Individual and average speech perception scores and RMS errors for the 5 conditions Speech perception
RMS errors (degrees)
Age
PTA (dB)
FMS (%)
HSM (%)
Combined
Bimodal
Bilateral Acoustic
HA Ipsi
HA Contra
68 74 26 65 46 73 66 53 67 50 51 20 69 58 68 58 70 26 56 17
92 90 75 48 97 43 67 60 72 56 63 84 86 68 43 75 70 75 70 16
25 40 50 60 70 65 70 85 20 70 75 65 55 45 82 20 65 48 56 20
71 79 100 84 93 92 60 100 36 100 97 92 79 84 100 66 78 48 81 19
83 (26) 81 (32) 47 (25) 53 (27) 63 (34) 88 (48) 100 (53) 65 (47) 74 (37) 74 (55) 68 (37) 53 (22) 75 (38) 67 (43) 80 (23) 60 (26) 66 (27) 71 (43) 70 13
77 (23) 84 (17) 73 (39) 104 (43) 84 (35) 93 (51) 105 (53) 91 (46) 70 (36) 80 (60) 71 (35) 67 (34) 54 (26) 85 (43) 59 (26) 91 (38) 59 (25) 83 (30) 79 15
76 (26) 96 (32) 58 (30) 78 (44) 63 (32) 103 (54) 85 (29) 71 (34) 73 (37) 71 (50) 75 (44) 80 (41) 85 (35) 76 (43) 105 (49) 63 (27) 79 (30) 78 (34) 79 13
96 (39) 104 (0) 74 (34) 93 (49) 104 (56) 103 (53) 99 (47) 81 (39) 73 (32) 76 (36) 99 (51) 101 (51) 86 (43) 90 (46) 98 (45) 104 (45) 93 (38) 98 (43) 93 11
94 (52) 97 (29) 99 (36) 90 (48) 103 (55) 103 (56) 100 (53) 88 (38) 96 (32) 84 (49) 105 (56) 95 (38) 103 (49) 97 (43) 99 (42) 97 (41) 90 (37) 98 (41) 97 6
The subjects’ most recent speech perception scores are presented. Standard deviations are presented within brackets. Contra indicates contralateral; FMS, Freiburger Monosyllabic word test; HA, hearing aid; HL, Hearing loss; HSM, Hochmair-Schulz-Moser sentence test; Ipsi, ipsilateral; PTA, pure-tone average (125Y1,000 Hz); SD, standard deviation. Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SOUND LOCALIZATION WITH A HYBRID CI
1437
scores in the combined or bilateral acoustic conditions and were able to take advantage of the additional acoustic information provided by the hybrid CI.
FIG. 4. Mean RMS errors in degrees for the 5 different conditions. Error bars indicate 95% confidence interval. Shaded bars indicate the monaural conditions. Average RMS error for normal hearing on this sound localization test is 20 degrees, measured binaurally. Chance RMS error is 104 degrees. *p G 0.05; NH: normal hearing.
Source Localization RMS errors in the 3 binaural conditions as a function of all 12 loudspeakers are presented in Figure 5A. Repeated measures analyses identified a significant effect of condition for the loudspeakers at 0 degree (F2,34 = 4.027; p = 0.027), 30 degrees (F2,34 = 7.659; p = 0.002), 60 degrees (F2,34 = 11.685; p = 0.001), and 90 degrees (F2,34 = 5.342; p = 0.022). Bonferroni post hoc analyses revealed a significantly better performance in the combined condition for the loudspeakers at 0 degree (p = 0.016), 30 degrees (p = 0.003), 60 degrees (p G 0.001), and 90 degrees (p = 0.003) as compared with the bilateral acoustic condition. In Figure 5B. the mean RMS errors in both monaural conditions are presented for all 18 subjects. Means and standard deviations for all loudspeaker positions are presented in Table 3. When both monaural conditions for the 13 subjects, with a conventional contralateral HA were compared, no significant differences were present between the contralateral side with the conventional HAs and the ipsilateral side.
(p = 0.55) and the combined and bilateral acoustic condition (p = 0.07). The individual results reveal extensive interindividual RMS error differences, for example, in the combined condition a range in RMS error between 47 and 100 degrees was observed (Table 2). Figure 4 displays the results of a subject with a moderate localization performance and a subject experiencing difficulties in determining the exact location of a sound. Fourteen patients obtained best
Factors of Influence Initially, 7 variables were identified as possible influencing factors: the subjects age at visit, age at implantation, duration of hearing loss, mean low-frequency hearing level, filter bank settings, and performance on monosyllabic word or sentence tests in quiet. As age at implantation was highly correlated with age at visit (r = 0.999, p G 0.001), this factor was not taken into account.
FIG. 5. A, Mean RMS errors for the binaural conditions as a function of the 12 loudspeaker positions, with 0 degree corresponding to loudspeaker 12. Scores closer to the center indicate a better performance. To correct for different sides of implantation, the data are converted in such a way that it implies that all subjects have worn their hybrid CI on one side, the ‘‘Hybrid-L side.’’ Asterisks denote a significantly better performance in the combined condition as compared to the bilateral acoustic condition (*p G 0.05; **p G 0.01). B, Mean RMS errors for both monaural conditions as a function of the different loudspeaker positions. Scores closer to the center indicate a better performance. To correct for different sides of implantation, the data are converted in such a way that it implies that all subjects wore their hybrid CI on one side, the ‘‘Hybrid-L side.’’ Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
1438
M. J. W. LAMMERS ET AL. TABLE 3.
12 (0 1 (30 2 (60 3 (90 4 (120 5 (150 6 (180 degree) degrees) degrees) degrees) degrees) degrees) degrees)
Condition Combined
RMS errors (degrees) for each loudspeaker position
Mean SD Bimodal Mean SD Bilateral Mean acoustic SD Ipsilateral Mean acoustic SD Contralateral Mean acoustic SD
53 31 69 31 74 37 99 31 104 37
38 23 56 26 68 38 62 30 115 28
37 19 52 29 79 35 50 27 127 31
41 27 55 32 69 35 34 29 117 30
65 38 80 31 80 42 52 30 116 33
98 24 97 37 110 40 76 29 107 42
101 40 107 44 101 33 82 35 87 37
7 (j150 degrees)
8 (j120 degrees)
9 (j90 degrees)
10 (j60 degrees)
11 (j30 degrees)
69 31 81 40 59 27 80 37 65 26
48 35 60 42 43 24 99 35 45 12
56 38 52 44 44 25 124 35 43 34
64 43 69 41 49 29 133 31 47 34
60 39 67 32 60 36 109 34 63 31
SD indicates standard deviation.
Results of the 6 analyses indicated a significant positive correlation of RMS error in the combined condition with age at visit (r = 0.597, p G 0.01), indicating poorer performance with increasing age (Fig. 6). The other factors were not found to be significant predictors of sound localization.
DISCUSSION These results demonstrate that the best localization abilities of hybrid CI users are expected with the fully aided combined condition, in which their hybrid CI is supported with their contralateral acoustic hearing. Although no significant differences were found between the binaural conditions (combined, bimodal, and bilateral acoustic), almost all subjects obtained their highest scores in the combined or bimodal condition, indicating the additional effect of the cochlear implant. This is supported by the significantly better performance in the combined condition for the loudspeakers at 0, 30, 60, and 90 degrees as compared with the bilateral acoustic condition. When discussing the different conditions, it should be noted that all subjects were not accustomed using any other condition than their daily used combined condition. Furthermore, subjects had little time to adapt to the other conditions tested, and the devices were not balanced before testing with the different listening conditions. Differences in outcome between the conditions could therefore be partly described by a negative effect because of the unfamiliarity with the other conditions. Longer adaptation times to all conditions and careful balancing procedures were, because of time constraints, not feasible in this study. Individual results of all subjects show an extensive variability between subjects and test conditions. These differences could, at least partially, be described by the significant relation between RMS error and age. One explanation for this correlation could be that older subjects experience more difficulties with this localization test. Because the objective of this study was not to determine the influence of several factors on sound localization abilities, it cannot be ruled out that other factors, such as residual low-frequency hearing and device characteristics, although not significantly correlated to performance in this study group, did not contribute to the extensive variability.
As shown in Figure 5, in all binaural conditions, front-back confusions are present, resulting in reduced performances for especially the loudspeakers in the back and to a lesser extent directly in the front. Sound sources in front of the subject on the CI side, were, however, better identified in the combined condition compared with the bilateral acoustic condition. Previous sound localization tests with 6 hybrid CI users in the same test setup produced comparable RMS errors for the 3 binaural conditions (12). Unlike the results of the present study, a significant difference between the combined and bimodal conditions was present in these 6 subjects. The extensive interindividual differences in combination with a small study population could have contributed to this different outcome. Noticeable were the equivalent results between the preoperative and postoperative bilateral acoustic conditions found in these 6 hybrid subjects, indicating that the postoperative bilateral acoustic condition is comparable to the subjects’ preoperative performance. Dunn et al. (8) demonstrated that users of a hybrid CI with a short electrode array achieved significantly better sound localization scores in the combined and bilateral hearing aid
FIG. 6. Individual RMS errors in the combined condition (in degrees) as a function of age. The straight line represents a linear regression fit through the data.
Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SOUND LOCALIZATION WITH A HYBRID CI conditions compared with the bimodal condition. In contrast to the subjects in our study, their results show that all subjects could integrate the bilateral low-frequency acoustic information. A possible explanation might be that the subjects in their study had lower pure tone audiogram thresholds for the frequencies 500, 750, and 1,000 Hz. Furthermore, the use of overlapping filter bank settings in 6 subjects might have induced interference of ITD cues, resulting in inferior localization ability in the combined condition. In an earlier unpublished study by our group, unilateral conventional CI users with contralateral hearing aids were tested in the same localization test. One year after implantation average RMS errors of 75 degrees were obtained. These results were not significantly different to the results with the hybrid users in all binaural conditions. The comparable results between the conventional and the hybrid CI might imply that the CI users tested in this setup mainly rely on the head shadow effect and are hardly able to integrate the bilateral acoustic hearing to aid them in sound localization. Because participants already experienced large difficulties with a fixed sound intensity, we did not use a roving intensity and were therefore not able to determine the extent and the use of the head shadow effect. The fact that the results in the bimodal condition are similar to the results obtained with the conventional CI indicates that sound localization abilities with the shorter array are equivalent to the conventional, longer electrode arrays. This is especially important for users who lose their residual hearing or users who are unable to use their residual hearing. Comparable findings of equivalence between the longer conventional and shorter electrodes have been reported for speech perception tests in quiet and noise (7). In this study, we have performed sound localization tests to assess the localization abilities of hybrid CI users in various everyday listening conditions. Because we used a broadband stimulus containing both ITD and ILD cues, future studies might address the influence of ITD and ILD cues separately by using band pass filter noise signals in conjunction with balanced hybrid CIs and hearing aids. Because participants already experienced large difficulties with a fixed sound intensity, we did not use a roving intensity. Because of the fixed intensity level, participants could have used the head shadow effect as a useful cue. CONCLUSION Results from this study indicate that binaural stimulation with a unilateral hybrid CI and contralateral acoustic stimulation offers significant benefits in sound localization. Best
1439
results, for most subjects, were obtained with the routinely used combined fitting and are comparable to results obtained with a bimodal fitting with a conventional CI. The additional acoustic benefit, generated by the acoustic stimulation of the implanted ear, improved sound localization for a vast majority of subjects, but a significant improvement was absent. On the other hand, a marked improvement has been found by the addition of electrical stimulation, for especially the azimuths ipsilateral and to the front of the CI, implying that especially the cochlear implant is responsible for the improvement in sound localization. Acknowledgments: The authors thank all the recipients participating in this study for their support and Mark Schuessler and Henrike Schultrich for the assistance with the measurements.
REFERENCES 1. Gantz BJ, Hansen MR, Turner CW, et al. Hybrid 10 clinical trial: preliminary results. Audiol Neurootol 2009;14(Suppl):32Y8. 2. Lorens A, Polak M, Piotrowska A, et al. Outcomes of treatment of partial deafness with cochlear implantation: a DUET study. Laryngoscope 2008;118:288Y94. 3. Gstoettner WK, Helbig S, Maier N, et al. Ipsilateral electric acoustic stimulation of the auditory system: results of long-term hearing preservation. Audiol Neurootol 2006;11(Suppl):49Y56. 4. Kiefer J, Pok M, Adunka O, et al. Combined electric and acoustic stimulation of the auditory system: results of a clinical study. Audiol Neurootol 2005;10:134. 5. Lenarz T, Stover T, Buechner A, et al. Temporal bone results and hearing preservation with a new straight electrode. Audiol Neurootol 2006;11(Suppl):34Y41. 6. Havenith S, Lammers MJ, Tange RA, et al. Hearing preservation surgery: cochleostomy or round window approach? A systematic review. Otol Neurotol 2013;34:667Y74. 7. Lenarz T, Stover T, Buechner A, et al. Hearing conservation surgery using the Hybrid-L electrode. Results from the first clinical trial at the Medical University of Hannover. Audiol Neurootol 2009;14(Suppl): 22Y31. 8. Dunn CC, Perreau A, Gantz B, et al. Benefits of localization and speech perception with multiple noise sources in listeners with a short-electrode cochlear implant. J Am Acad Audiol 2010;21:44Y51. 9. Grantham DW, Ashmead DH, Ricketts TA, et al. Interaural time and level difference thresholds for acoustically presented signals in post-lingually deafened adults fitted with bilateral cochlear implants using CIS+ processing. Ear Hear 2008;29:33Y44. 10. Francart T, Brokx J, Wouters J. Sensitivity to interaural level difference and loudness growth with bilateral bimodal stimulation. Audiol Neurootol 2008;13:309Y19. 11. Bu¨chner A, Schussler M, Battmer RD, et al. Impact of lowfrequency hearing. Audiol Neurootol 2009;14(Suppl):8Y13. 12. Bo¨hm M, Pesch J, Battmer RD, et al. Lokalisationfa¨higkeit bei Hybrid L Patienten German Medical Science Web site, 2007. 13. Laszig R, Aschendorff A, Stecker M, et al. Benefits of bilateral electrical stimulation with the nucleus cochlear implant in adults: 6-month postoperative results. Otol Neurotol 2004;25:958Y68.
Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
Sound Localization Abilities of Unilateral Hybrid Cochlear Implant Users With Bilateral Low-Frequency Hearing *†Marc J. W. Lammers, ‡Thomas Lenarz, *†Gijsbert A. van Zanten, *†Wilko Grolman, and §Andreas Buechner *Department of OtorhinolaryngologyYHead and Neck Surgery, ÞBrain Center Rudolf Magnus, University Medical Center Utrecht, The Netherlands; þCluster of Excellence ‘‘Hearing4All,’’ Department of Otolaryngology, and §Cluster of Excellence ‘‘Hearing4All,’’ German Hearing Center, Hannover Medical School, Germany
Objective: To assess the sound localization abilities of subjects unilaterally implanted with a hybrid cochlear implant in different sound localization conditions. Study Design: A prospective, single-subject repeated measures design was performed to assess the sound localization abilities in 5 different listening conditions: combined (hybrid cochlear implant and contralateral acoustic hearing), bimodal, bilateral acoustic, ipsilateral acoustic, and contralateral acoustic. Setting: Tertiary referral center. Patients: Eighteen subjects with bilateral low-frequency residual hearing, implanted with a hybrid cochlear implant, were enrolled in this study. Main Outcome Measures: Postoperative sound localization in 5 different listening conditions. Results: Average group results showed a significant improvement in sound localization for all binaural hearing conditions
(combined, bimodal, and bilateral acoustic) over monaural conditions (ipsilateral acoustic and contralateral acoustic). Subjects performed significantly better in the combined condition compared with the bimodal and bilateral acoustic conditions when sound was presented from the front and the side of their cochlear implant. Conclusion: Best results, for most subjects, were obtained with the routinely used combined fitting. The additional acoustic stimulation of the implanted ear did not significantly improve sound localization. However, a marked improvement has been found by the addition of electrical stimulation, for especially the azimuths ipsilateral and to the front of the cochlear implant, implying the importance of the cochlear implant in sound localization. Key Words: Cochlear implantVElectro-acoustic stimulationVHearing aidVResidual hearingVSound localization. Otol Neurotol 35:1433Y1439, 2014.
Combining electrical speech processing with lowfrequency residual hearing is increasingly becoming a more common strategy for patients with severe-to-profound high-frequency hearing loss who are unable to obtain satisfactory results with modern hearing aids. Preservation of the residual hearing in these subjects is essential for obtaining the best results. Accumulating evidence reveals that this might be possible by using shorter and less traumatic electrodes and that auditory performance is
comparable or even better than with a conventional cochlear implant (CI) (1Y6). Recent results with this hybrid CI have demonstrated a high probability of postoperative residual hearing preservation and improved auditory performance in the electroacoustic stimulation condition (7). This preserved residual hearing might be especially beneficial in complex auditory conditions, such as speech in background noise or sound localization (1,8). When localizing sounds, traditional CI users mainly rely on interaural level differences (ILD) (9,10). The additional low-frequency acoustic stimulation in patients with residual hearing could provide them with interaural time difference (ITD) cues, which might aid them in their sound localization (8). To date, little is known about the sound localization abilities of patients who have undergone hearing preservation surgery with cochlear implantation and who have 2 ears with sensitivity to low-frequency acoustic stimulation.
Address correspondence and reprint requests to Andreas Buechner, Ph.D., Department of Otolaryngology, Medizinische Hochschule Hannover, Carl Neuberg Strasse 1, DEY30625 Hannover, Germany; E-mail: [email protected] Financial disclosure: This research received no specific grants from any funding agency in the public, commercial, or not-for-profit sectors. Competing Interests: Wilko Grolman received unrestrictive research grants from Cochlear Ltd., Med-El GmbH, and Advanced Bionics. No competing interests declared by Marc Lammers, Thomas Lenarz, Gijsbert van Zanten, or Andreas Buechner.
1433
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
1434
M. J. W. LAMMERS ET AL.
One study reported on the sound localization abilities of subjects implanted with a short electrode and demonstrated that these subjects are able to use their residual hearing for sound localization (8). In this study, we evaluated the sound localization abilities of patients with bilateral residual hearing, who have been fitted with a unilateral hybrid cochlear implant.
(open squares), as well as the average preoperative audiograms for the implanted ear (filled circles), are displayed in Figure 1. A postoperative pure-tone average (PTA) was calculated using the threshold levels measured at 125, 250, 500, 750, and 1,000 Hz. If there was no response up to the limit of the audiometer, a nominal value of 130 dB was assigned for the purpose of calculation, indicating complete hearing loss.
MATERIALS AND METHODS
A prospective single-subject repeated measures design was performed in a cohort of 18 hybrid CI users. All subjects were tested in 5 different conditions to evaluate the localization abilities using the hybrid CI and their contralateral ear with or without hearing aids. Conditions were tested in random order and consisted of the following: combined, bimodal, bilateral acoustic only, ipsilateral acoustic only, and contralateral acoustic only. The combined condition referred to the hybrid CI and acoustic stimulation (AS) in the contralateral ear (CIHA + AS). In the bimodal condition only the CI of the hybrid device with contralateral acoustic stimulation was tested (CI + AS). The HA of the hybrid CI was removed and the implanted ear was plugged and covered with earmuffs to minimize acoustic hearing. In the bilateral acoustic condition, the CI was switched off, so that both ears were solely acoustically stimulated (HA + AS). In the ipsilateral acoustic condition, the contralateral ear was plugged and covered with earmuffs, whereas in the implanted ear, only the HA of the hybrid CI was switched on. In the contralateral acoustic condition, the hybrid CI was removed, and the implanted ear was plugged and covered with earmuffs.
Study Design Subjects Eighteen adults, with bilateral residual hearing and unilaterally implanted with the Cochlear Nucleus Hybrid-L device (Cochlear Corp., Australia), were enrolled in this study. Subject demographics of the subjects are shown in Table 1. Most subjects had experience with the sound localization test. Filter bank settings of the hybrid CI were set to either an overlapping map or a nonoverlapping map. In the overlapping map setting, the full frequency spectrum of approximately 200 to 7,000 Hz is transmitted through the cochlear implant, whereas in the nonoverlapping map setting a reduced frequency range, depending on the residual hearing of the individual subject, is transmitted (11). Twelve subjects used the nonoverlapping filter bank settings, and 13 subjects used a contralateral hearing aid, as indicated in Table 1. The contralateral hearing aids were fitted by hearing aid acousticians and were checked prior to the test procedure for proper functioning. The fitting itself was not changed before testing and their daily settings were used. Cochlear Hybrid Prescription was used for the hearing aids of the hybrid CI. Loudness balancing was done during the fitting of the CI, i.e. the loudness of the CI was adapted such that the perceived loudness in live voice was equal between the left and right ear. The group average postoperative pure tone audiogram for the implanted ear (solid triangles) and the contralateral ear TABLE 1. Age at Experience Subject Age implantation with CI no. (yr) (yr) (mo) 1 2
67,9 74,0
67,7 73,1
1 9
3 4 5 6 7 8 9
26,2 65,3 45,7 73,4 65,7 52,5 67,1
23,3 63,1 44,7 72,8 62,3 51,3 65,7
34 25 11 6 40 13 16
10 11 12 13 14 15 16 17
49,9 50,6 20,0 69,4 58,4 68,3 57,6 70,3
49,0 49,8 17,8 67,7 56,0 66,4 56,9 68,9
10 8 25 20 27 20 6 14
18 Mean
25,8 56,0
24,6 54,5
13 17
Etiology Unknown Sudden deafness Preterm birth Ototoxic Trauma Genetic Unknown Ototoxic Sudden deafness Unknown Unknown Unknown Unknown Rubella Otitis Media Genetic Sudden deafness Unknown
Sound Localization Sound localization tests were performed in a sound-treated booth (12 m2) with 12 loudspeakers spaced equidistantly at 30-degree intervals spanning 360 degrees, located approximately 1 m above the floor, and numbered from 1 to 12 in a clockwise
Subject demographics
Duration of HA Duration of HL Duration of HA Contralateral use contralateral in implanted use in implanted ear ear (yr) ear (yr) ear (yr)
Filter bank setting
LF (Hz)
Overlapping Overlapping
188 188
HA HA
11 9
16 21
11 9
HA No HA HA HA No HA HA No HA
19 40a 22 5 20a 13 19a
24 40 45 8 22 20 15
19 40 36 6 15 13 15
Nonoverlapping 813 Nonoverlapping 688 Nonoverlapping 438 Nonoverlapping 1188 Overlapping 188 Nonoverlapping 438 Nonoverlapping 813
No HA HA HA HA HA HA HA HA
5a 19 17 15 24 53 43 4
1 50 20 15 33 68 57 4
19 17 15 24 53 37 4
Nonoverlapping 688 Nonoverlapping 1688 Overlapping 188 Nonoverlapping 438 Overlapping 188 Nonoverlapping 1188 Nonoverlapping 688 Nonoverlapping 563
No HA
-a 20
25 27
21
a
Overlapping
Subject was not using hearing aids (HAs) at the time of testing. CI indicates cochlear implant; HA, hearing aid; HL, hearing loss; LF, lowest frequency transmitted through the cochlear implant.
Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
188
SOUND LOCALIZATION WITH A HYBRID CI
1435
lateral HA, nonoverlapping map users with contralateral HA, overlapping map users without contralateral HA, and nonoverlapping map users without HA) were analyzed with KruskalWallis and Mann-Whitney U tests. If group differences were not statistically significant, pooling of the data was deemed possible. Differences between the 5 different conditions and the 12 sound source localizations were analyzed using repeated measures ANOVAs. Significant main effects ( p G 0.05) were followed-up with Bonferroni post hoc tests, and the Greenhouse-Geisser correction was applied to compensate for violations of the sphericity assumption. Linear regression analyses were performed to evaluate the influence of variables, such as degree of hearing loss and age at visit on sound localization performance.
RESULTS
FIG. 1. Group average pure tone audiograms (n = 18). The subjects’ most recent postoperative audiogram for the implanted ear (solid triangles) and preoperative audiogram for the implanted ear (filled circles) and contralateral ear (open squares) are presented.
sequence (Fig. 2). The subject was seated in the center of the circular array of loudspeakers, facing the number 12 loudspeaker, at a distance of 1.2 meters. The stimulus consisted of one short sentence from the Hochmair-Schulz-Moser (HSM) sentence test and was presented by one of the 12 loudspeakers, in a random order. The sound level was fixed at a level of 65 dB SPL. No sound intensity roving was applied. This stimulus configuration was used to compare our results with previous sound localization tests with the hybrid CI and conventional CI users, conducted in our center (12,13). Every test consisted of four presentations by all 12 loudspeakers. During the test, patients were instructed to keep their heads steadily directed toward loudspeaker 12. Head movements were visually monitored by the investigator. Subjects were asked to verbally identify the number corresponding to the loudspeaker from which the sentence was presented. The response errors in all test conditions were calculated and converted to degrees and root mean square (RMS) errors. Before the start of the test, subjects were given a short sample, by presenting the sentence from loudspeaker 12, 3, 6, and 9 (0, 90, 180, and j90 degrees). During the test, only one repeat was allowed if subjects could not hear the first stimulus. If a subject failed to complete the test, the chance RMS error of 104 degrees was awarded. No feedback was provided. All subjects used their routinely used filter bank setting and were not allowed to change their map settings. Because of time constraints, the hybrid CI and, if applicable, the contralateral hearing aid were not loudness balanced before the sound localization test. Neither were differences in levels in the monaural and binaural conditions corrected for binaural summation effects. The present study was performed according to the principles of the Declaration of Helsinki and received the approval of the ethical committee of the Medical University of Hannover.
Statistical Analyses Statistical analyses were completed using SPSS version 20.0 software. Group differences between the different patient groups included in this study (i.e., overlapping map users with contra-
Figure 3 displays the results of a subject with a moderate localization performance (upper row) and a subject experiencing difficulties in determining the exact location of a sound (lower row). Scores did not significantly differ between hearing aid and nonhearing aid users for all 5 conditions (Mann-Whitney U tests: p 9 0.22). Results of subjects using the overlapping map did not significantly differ from the results of users with the nonoverlapping maps (Mann-Whitney U tests: p 9 0.21). Because the average results of the 4 groups (overlapping with or without HA and nonoverlapping with or without HA) did not differ significantly, the data obtained from all 18 subjects was pooled. Individual and mean RMS errors for all 18 subjects in the 5 different conditions are presented in Table 2 and Figure 4. Repeated measure analyses indicated that significant differences between the different conditions were present (F4,68 = 19.081; p G 0.001). Bonferroni post hoc analyses revealed a significantly better performance in all
FIG. 2. Sound localization test setup. The speakers were spaced equidistantly at 30-degree intervals spanning 360 degrees at 1.2 meters from the subject. Subject was seated in the center, facing loudspeaker 12. Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
1436
M. J. W. LAMMERS ET AL.
FIG. 3. Example of localization results (in degrees) of 2 subjects (Subjects 3 and 6) for the 3 binaural conditions. A subject with moderate performance is presented on top, and a subject with poor performance in the bottom row. Means and standard deviations (T error bars) of all responses for each of the 12 loudspeaker positions are shown. The diagonal line represents perfect performance. Because of the 360degree test setup, response angles can vary from þ180 to j180 degrees around the source angle, that is, responses for a source angle of 150 degrees can vary from j30 to 330 degrees; 0 to 180 degrees refers to the Hybrid-L side.
binaural conditions (combined, bimodal, and bilateral acoustic) as compared with the monaural conditions (ipsilateral acoustic and contralateral acoustic) (p G 0.05). All conditions performed significantly better than the chance TABLE 2. Subject no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Mean SD
RMS error of 104 degrees (p G 0.01). There were no significant differences found between both monaural conditions. A nonsignificant RMS error difference of 9 degrees was found between the combined and the bimodal condition
Individual and average speech perception scores and RMS errors for the 5 conditions Speech perception
RMS errors (degrees)
Age
PTA (dB)
FMS (%)
HSM (%)
Combined
Bimodal
Bilateral Acoustic
HA Ipsi
HA Contra
68 74 26 65 46 73 66 53 67 50 51 20 69 58 68 58 70 26 56 17
92 90 75 48 97 43 67 60 72 56 63 84 86 68 43 75 70 75 70 16
25 40 50 60 70 65 70 85 20 70 75 65 55 45 82 20 65 48 56 20
71 79 100 84 93 92 60 100 36 100 97 92 79 84 100 66 78 48 81 19
83 (26) 81 (32) 47 (25) 53 (27) 63 (34) 88 (48) 100 (53) 65 (47) 74 (37) 74 (55) 68 (37) 53 (22) 75 (38) 67 (43) 80 (23) 60 (26) 66 (27) 71 (43) 70 13
77 (23) 84 (17) 73 (39) 104 (43) 84 (35) 93 (51) 105 (53) 91 (46) 70 (36) 80 (60) 71 (35) 67 (34) 54 (26) 85 (43) 59 (26) 91 (38) 59 (25) 83 (30) 79 15
76 (26) 96 (32) 58 (30) 78 (44) 63 (32) 103 (54) 85 (29) 71 (34) 73 (37) 71 (50) 75 (44) 80 (41) 85 (35) 76 (43) 105 (49) 63 (27) 79 (30) 78 (34) 79 13
96 (39) 104 (0) 74 (34) 93 (49) 104 (56) 103 (53) 99 (47) 81 (39) 73 (32) 76 (36) 99 (51) 101 (51) 86 (43) 90 (46) 98 (45) 104 (45) 93 (38) 98 (43) 93 11
94 (52) 97 (29) 99 (36) 90 (48) 103 (55) 103 (56) 100 (53) 88 (38) 96 (32) 84 (49) 105 (56) 95 (38) 103 (49) 97 (43) 99 (42) 97 (41) 90 (37) 98 (41) 97 6
The subjects’ most recent speech perception scores are presented. Standard deviations are presented within brackets. Contra indicates contralateral; FMS, Freiburger Monosyllabic word test; HA, hearing aid; HL, Hearing loss; HSM, Hochmair-Schulz-Moser sentence test; Ipsi, ipsilateral; PTA, pure-tone average (125Y1,000 Hz); SD, standard deviation. Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SOUND LOCALIZATION WITH A HYBRID CI
1437
scores in the combined or bilateral acoustic conditions and were able to take advantage of the additional acoustic information provided by the hybrid CI.
FIG. 4. Mean RMS errors in degrees for the 5 different conditions. Error bars indicate 95% confidence interval. Shaded bars indicate the monaural conditions. Average RMS error for normal hearing on this sound localization test is 20 degrees, measured binaurally. Chance RMS error is 104 degrees. *p G 0.05; NH: normal hearing.
Source Localization RMS errors in the 3 binaural conditions as a function of all 12 loudspeakers are presented in Figure 5A. Repeated measures analyses identified a significant effect of condition for the loudspeakers at 0 degree (F2,34 = 4.027; p = 0.027), 30 degrees (F2,34 = 7.659; p = 0.002), 60 degrees (F2,34 = 11.685; p = 0.001), and 90 degrees (F2,34 = 5.342; p = 0.022). Bonferroni post hoc analyses revealed a significantly better performance in the combined condition for the loudspeakers at 0 degree (p = 0.016), 30 degrees (p = 0.003), 60 degrees (p G 0.001), and 90 degrees (p = 0.003) as compared with the bilateral acoustic condition. In Figure 5B. the mean RMS errors in both monaural conditions are presented for all 18 subjects. Means and standard deviations for all loudspeaker positions are presented in Table 3. When both monaural conditions for the 13 subjects, with a conventional contralateral HA were compared, no significant differences were present between the contralateral side with the conventional HAs and the ipsilateral side.
(p = 0.55) and the combined and bilateral acoustic condition (p = 0.07). The individual results reveal extensive interindividual RMS error differences, for example, in the combined condition a range in RMS error between 47 and 100 degrees was observed (Table 2). Figure 4 displays the results of a subject with a moderate localization performance and a subject experiencing difficulties in determining the exact location of a sound. Fourteen patients obtained best
Factors of Influence Initially, 7 variables were identified as possible influencing factors: the subjects age at visit, age at implantation, duration of hearing loss, mean low-frequency hearing level, filter bank settings, and performance on monosyllabic word or sentence tests in quiet. As age at implantation was highly correlated with age at visit (r = 0.999, p G 0.001), this factor was not taken into account.
FIG. 5. A, Mean RMS errors for the binaural conditions as a function of the 12 loudspeaker positions, with 0 degree corresponding to loudspeaker 12. Scores closer to the center indicate a better performance. To correct for different sides of implantation, the data are converted in such a way that it implies that all subjects have worn their hybrid CI on one side, the ‘‘Hybrid-L side.’’ Asterisks denote a significantly better performance in the combined condition as compared to the bilateral acoustic condition (*p G 0.05; **p G 0.01). B, Mean RMS errors for both monaural conditions as a function of the different loudspeaker positions. Scores closer to the center indicate a better performance. To correct for different sides of implantation, the data are converted in such a way that it implies that all subjects wore their hybrid CI on one side, the ‘‘Hybrid-L side.’’ Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
1438
M. J. W. LAMMERS ET AL. TABLE 3.
12 (0 1 (30 2 (60 3 (90 4 (120 5 (150 6 (180 degree) degrees) degrees) degrees) degrees) degrees) degrees)
Condition Combined
RMS errors (degrees) for each loudspeaker position
Mean SD Bimodal Mean SD Bilateral Mean acoustic SD Ipsilateral Mean acoustic SD Contralateral Mean acoustic SD
53 31 69 31 74 37 99 31 104 37
38 23 56 26 68 38 62 30 115 28
37 19 52 29 79 35 50 27 127 31
41 27 55 32 69 35 34 29 117 30
65 38 80 31 80 42 52 30 116 33
98 24 97 37 110 40 76 29 107 42
101 40 107 44 101 33 82 35 87 37
7 (j150 degrees)
8 (j120 degrees)
9 (j90 degrees)
10 (j60 degrees)
11 (j30 degrees)
69 31 81 40 59 27 80 37 65 26
48 35 60 42 43 24 99 35 45 12
56 38 52 44 44 25 124 35 43 34
64 43 69 41 49 29 133 31 47 34
60 39 67 32 60 36 109 34 63 31
SD indicates standard deviation.
Results of the 6 analyses indicated a significant positive correlation of RMS error in the combined condition with age at visit (r = 0.597, p G 0.01), indicating poorer performance with increasing age (Fig. 6). The other factors were not found to be significant predictors of sound localization.
DISCUSSION These results demonstrate that the best localization abilities of hybrid CI users are expected with the fully aided combined condition, in which their hybrid CI is supported with their contralateral acoustic hearing. Although no significant differences were found between the binaural conditions (combined, bimodal, and bilateral acoustic), almost all subjects obtained their highest scores in the combined or bimodal condition, indicating the additional effect of the cochlear implant. This is supported by the significantly better performance in the combined condition for the loudspeakers at 0, 30, 60, and 90 degrees as compared with the bilateral acoustic condition. When discussing the different conditions, it should be noted that all subjects were not accustomed using any other condition than their daily used combined condition. Furthermore, subjects had little time to adapt to the other conditions tested, and the devices were not balanced before testing with the different listening conditions. Differences in outcome between the conditions could therefore be partly described by a negative effect because of the unfamiliarity with the other conditions. Longer adaptation times to all conditions and careful balancing procedures were, because of time constraints, not feasible in this study. Individual results of all subjects show an extensive variability between subjects and test conditions. These differences could, at least partially, be described by the significant relation between RMS error and age. One explanation for this correlation could be that older subjects experience more difficulties with this localization test. Because the objective of this study was not to determine the influence of several factors on sound localization abilities, it cannot be ruled out that other factors, such as residual low-frequency hearing and device characteristics, although not significantly correlated to performance in this study group, did not contribute to the extensive variability.
As shown in Figure 5, in all binaural conditions, front-back confusions are present, resulting in reduced performances for especially the loudspeakers in the back and to a lesser extent directly in the front. Sound sources in front of the subject on the CI side, were, however, better identified in the combined condition compared with the bilateral acoustic condition. Previous sound localization tests with 6 hybrid CI users in the same test setup produced comparable RMS errors for the 3 binaural conditions (12). Unlike the results of the present study, a significant difference between the combined and bimodal conditions was present in these 6 subjects. The extensive interindividual differences in combination with a small study population could have contributed to this different outcome. Noticeable were the equivalent results between the preoperative and postoperative bilateral acoustic conditions found in these 6 hybrid subjects, indicating that the postoperative bilateral acoustic condition is comparable to the subjects’ preoperative performance. Dunn et al. (8) demonstrated that users of a hybrid CI with a short electrode array achieved significantly better sound localization scores in the combined and bilateral hearing aid
FIG. 6. Individual RMS errors in the combined condition (in degrees) as a function of age. The straight line represents a linear regression fit through the data.
Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
SOUND LOCALIZATION WITH A HYBRID CI conditions compared with the bimodal condition. In contrast to the subjects in our study, their results show that all subjects could integrate the bilateral low-frequency acoustic information. A possible explanation might be that the subjects in their study had lower pure tone audiogram thresholds for the frequencies 500, 750, and 1,000 Hz. Furthermore, the use of overlapping filter bank settings in 6 subjects might have induced interference of ITD cues, resulting in inferior localization ability in the combined condition. In an earlier unpublished study by our group, unilateral conventional CI users with contralateral hearing aids were tested in the same localization test. One year after implantation average RMS errors of 75 degrees were obtained. These results were not significantly different to the results with the hybrid users in all binaural conditions. The comparable results between the conventional and the hybrid CI might imply that the CI users tested in this setup mainly rely on the head shadow effect and are hardly able to integrate the bilateral acoustic hearing to aid them in sound localization. Because participants already experienced large difficulties with a fixed sound intensity, we did not use a roving intensity and were therefore not able to determine the extent and the use of the head shadow effect. The fact that the results in the bimodal condition are similar to the results obtained with the conventional CI indicates that sound localization abilities with the shorter array are equivalent to the conventional, longer electrode arrays. This is especially important for users who lose their residual hearing or users who are unable to use their residual hearing. Comparable findings of equivalence between the longer conventional and shorter electrodes have been reported for speech perception tests in quiet and noise (7). In this study, we have performed sound localization tests to assess the localization abilities of hybrid CI users in various everyday listening conditions. Because we used a broadband stimulus containing both ITD and ILD cues, future studies might address the influence of ITD and ILD cues separately by using band pass filter noise signals in conjunction with balanced hybrid CIs and hearing aids. Because participants already experienced large difficulties with a fixed sound intensity, we did not use a roving intensity. Because of the fixed intensity level, participants could have used the head shadow effect as a useful cue. CONCLUSION Results from this study indicate that binaural stimulation with a unilateral hybrid CI and contralateral acoustic stimulation offers significant benefits in sound localization. Best
1439
results, for most subjects, were obtained with the routinely used combined fitting and are comparable to results obtained with a bimodal fitting with a conventional CI. The additional acoustic benefit, generated by the acoustic stimulation of the implanted ear, improved sound localization for a vast majority of subjects, but a significant improvement was absent. On the other hand, a marked improvement has been found by the addition of electrical stimulation, for especially the azimuths ipsilateral and to the front of the CI, implying that especially the cochlear implant is responsible for the improvement in sound localization. Acknowledgments: The authors thank all the recipients participating in this study for their support and Mark Schuessler and Henrike Schultrich for the assistance with the measurements.
REFERENCES 1. Gantz BJ, Hansen MR, Turner CW, et al. Hybrid 10 clinical trial: preliminary results. Audiol Neurootol 2009;14(Suppl):32Y8. 2. Lorens A, Polak M, Piotrowska A, et al. Outcomes of treatment of partial deafness with cochlear implantation: a DUET study. Laryngoscope 2008;118:288Y94. 3. Gstoettner WK, Helbig S, Maier N, et al. Ipsilateral electric acoustic stimulation of the auditory system: results of long-term hearing preservation. Audiol Neurootol 2006;11(Suppl):49Y56. 4. Kiefer J, Pok M, Adunka O, et al. Combined electric and acoustic stimulation of the auditory system: results of a clinical study. Audiol Neurootol 2005;10:134. 5. Lenarz T, Stover T, Buechner A, et al. Temporal bone results and hearing preservation with a new straight electrode. Audiol Neurootol 2006;11(Suppl):34Y41. 6. Havenith S, Lammers MJ, Tange RA, et al. Hearing preservation surgery: cochleostomy or round window approach? A systematic review. Otol Neurotol 2013;34:667Y74. 7. Lenarz T, Stover T, Buechner A, et al. Hearing conservation surgery using the Hybrid-L electrode. Results from the first clinical trial at the Medical University of Hannover. Audiol Neurootol 2009;14(Suppl): 22Y31. 8. Dunn CC, Perreau A, Gantz B, et al. Benefits of localization and speech perception with multiple noise sources in listeners with a short-electrode cochlear implant. J Am Acad Audiol 2010;21:44Y51. 9. Grantham DW, Ashmead DH, Ricketts TA, et al. Interaural time and level difference thresholds for acoustically presented signals in post-lingually deafened adults fitted with bilateral cochlear implants using CIS+ processing. Ear Hear 2008;29:33Y44. 10. Francart T, Brokx J, Wouters J. Sensitivity to interaural level difference and loudness growth with bilateral bimodal stimulation. Audiol Neurootol 2008;13:309Y19. 11. Bu¨chner A, Schussler M, Battmer RD, et al. Impact of lowfrequency hearing. Audiol Neurootol 2009;14(Suppl):8Y13. 12. Bo¨hm M, Pesch J, Battmer RD, et al. Lokalisationfa¨higkeit bei Hybrid L Patienten German Medical Science Web site, 2007. 13. Laszig R, Aschendorff A, Stecker M, et al. Benefits of bilateral electrical stimulation with the nucleus cochlear implant in adults: 6-month postoperative results. Otol Neurotol 2004;25:958Y68.
Otology & Neurotology, Vol. 35, No. 8, 2014
Copyright © 2014 Otology & Neurotology, Inc. Unauthorized reproduction of this article is prohibited.
