Cochlear Implant and Hearing Aid: Objective Measures of Binaural Benefit

Abstract

Cochlear implants (CI) improve hearing for the severely hearing impaired. With an extension of implantation candidacy, today many CI listeners use a hearing aid on their contralateral ear, referred to as bimodal listening. It is uncertain, however, whether the brains of bimodal listeners can combine the electrical and acoustical sound information and how much CI experience is needed to achieve an improved performance with bimodal listening. Patients with bilateral sensorineural hearing loss undergoing implant surgery were tested in their ability to understand speech in quiet and in noise, before and again 3 and 6 months after provision of a CI. Results of these bimodal listeners were compared to age-matched, normal hearing controls (NH). The benefit of adding a contralateral hearing aid was calculated in terms of head shadow, binaural summation, binaural squelch, and spatial release from masking from the results of a sentence recognition test. Beyond that, bimodal benefit was estimated from the difference in amplitudes and latencies of the N1, P2, and N2 potentials of the brains' auditory evoked response (AEP) toward speech. Data of fifteen participants contributed to the results. CI provision resulted in significant improvement of speech recognition with the CI ear, and in taking advantage of the head shadow effect for understanding speech in noise. Some amount of binaural processing was suggested by a positive binaural summation effect 6 month post-implantation that correlated significantly with symmetry of pure tone thresholds. Moreover, a significant negative correlation existed between binaural summation and latency of the P2 potential. With CI experience, morphology of the N1 and P2 potentials in the AEP response approximated that of NH, whereas, N2 remained different. Significant AEP differences between monaural and binaural processing were shown for NH and for bimodal listeners 6 month post-implantation. Although the grand-averaged difference in N1 amplitude between monaural and binaural listening was similar for NH and the bimodal group, source localization showed group-dependent differences in auditory and speech-relevant cortex, suggesting different processing in the bimodal listeners.

Keywords: auditory evoked potentials; auditory rehabilitation; bimodal benefit; cochlear implant; electroencephalography; hearing aid; source localization; speech recognition.

FIGURE 1

Improvement of hearing with the CI ear between T3 and T4. Significant improvements are seen for various test constellations in quiet (A) and background noise (B). In the FBE higher values (%-correct) indicate better performance, whereas in all OlSa tests, lower values indicate better performance. Group means with their standard errors are shown (**p < 0.002, *p < 0.01).

FIGURE 2

Speech perception with FBE and OlSa tests (A–F) and assessed in the AEP experiment (G). Higher values signal better speech recognition in the FBE and AEP conditions, whereas lower values indicate better speech recognition in OlSa tests. Note the reversed vertical scale in (B). Perception is best in NH listeners (A,B,G), worst shortly after CI provision (C,D,G) and improves with CI experience (E–G). Statistically significant differences between monaural listening with the CI or the designated CI ear in NH and binaural speech recognition was observed for several test conditions. Due to insufficient monaural hearing, monaural data at T2 are only available for the AEP condition, where a significant difference existed between monaural and binaural listening. While behavioral results from the AEP experiment at T3 and T4 did not evidence a significant difference between listening conditions, behavioral tests showed significantly better bimodal speech recognition at T3 for all test conditions and at T4 for the S0NCI condition. Regarding the spatial arrangement of speech and noise sources, the AEP condition is closest to S0NHA. Means and their standard error are shown (**p < 0.01, *p < 0.05, trends ⁺p < 0.1).

FIGURE 3

Grand averages for monaural, binaural listening conditions, and the difference binaural – monaural for the categories “words” (A–D) and “reversals” (E–H) of the CI (T2–T4) and NH group. (A–H) Time intervals with N1, P2, and N2 responses are shaded in different grays.

FIGURE 4

Quantitative AEP results: (A) mean amplitude and (B) area latency of the N1, (C) area latency of P2, and (D) mean amplitude of N2 for the categories “words”, “reversals”, and listening conditions monCI and binaural. (A–D) Means with their standard deviations are shown; significant differences between stimulus categories and listening conditions are indicated (**p < 0.001, *p < 0.05, and trends ⁺p < 0.1).

FIGURE 5

Spatial spread of monaural vs. binaural activation in CI listeners at T4 (A) and in NH (B) during the N1 interval. Only differences in regions bordering the surface or the midline of the cortex are visible in this illustration. For a complete list of areas with differential activation (see Table 6). Differences are more widespread in CI listeners compared to NH. Whereas in NH differential activation located to primary and secondary auditory cortex (BA41, 42), it pertained to auditory association cortex related to speech processing (BA21, 38, 39) and a region related with these areas (BA7) in the bimodal listeners. Darkening of the color scale indicates decreasing p values or higher significance.