Psychophysically based site selection coupled with dichotic stimulation improves speech recognition in noise with bilateral cochlear implants

Abstract

The ability to perceive important features of electrical stimulation varies across stimulation sites within a multichannel implant. The aim of this study was to optimize speech processor MAPs for bilateral implant users by identifying and removing sites with poor psychophysical performance. The psychophysical assessment involved amplitude-modulation detection with and without a masker, and a channel interaction measure quantified as the elevation in modulation detection thresholds in the presence of the masker. Three experimental MAPs were created on an individual-subject basis using data from one of the three psychophysical measures. These experimental MAPs improved the mean psychophysical acuity across the electrode array and provided additional advantages such as increasing spatial separations between electrodes and/or preserving frequency resolution. All 8 subjects showed improved speech recognition in noise with one or more experimental MAPs over their everyday-use clinical MAP. For most subjects, phoneme and sentence recognition in noise were significantly improved by a dichotic experimental MAP that provided better mean psychophysical acuity, a balanced distribution of selected stimulation sites, and preserved frequency resolution. The site-selection strategies serve as useful tools for evaluating the importance of psychophysical acuities needed for good speech recognition in implant users.

Figures

Figure 1

Psychophysical measures for all available stimulation sites in the left (open symbols) and right (filled symbols) ears. The first column shows means and ranges of modulation detection thresholds (MDTs) in quiet. The second column shows means and ranges of masked MDTs. The third column shows the amount of masking (masked MDTs minus MDTs in quiet). Correlation coefficients (r) for the across-site correlations between two ears on the psychophysical measure are shown in the upper right corner of each panel. Asterisks indicate statistically significant correlations between the two ears in the across-site patterns (p < 0.05). The ASMs of the left and right ears for the three measures are shown in the upper left corner of each panel. For each subject, the psychophysical measure chosen for site selection is indicated by text “MSS” shown at the lower right corner of the corresponding panel.

Figure 2

Schematic representation of site selection and frequency allocation for the three experimental MAPs. MDTs measured from S86 are shown above the implant diagrams. The filled symbols stand for the right ear, and the open ones stand for the left ear. The stimulation sites marked with squares were those that were turned off. In the implant diagrams, ellipses represent electrodes, with filled ones representing those that were turned off in the MAP. Open symbols represent electrodes that were used. Numbers in the open ellipses stand for the numbers designating the analysis bands assigned to the electrodes. In MAP A, for each stimulation site, the ear with better MDT was selected while the site on the opposite ear was turned off. In MAP B, the electrode array was divided into 5 segments. Within each segment, two sites with relatively poor MDT, each from a different ear, were turned off. In MAPs A and B, there were spectral “holes” in each ear at the stimulation sites that were turned off. However, these spectral contents were presented in the opposite ear. In MAP C, the same stimulation sites were turned off as in MAP B, but the frequencies of the electrode arrays in both implants were re-adjusted, so each implant contained 17 contiguous frequency bands with widened bandwidths.

Figure 3

Scatter plots showing the correlations between parings of the three psychophysical measures. The left columns show correlations of the across-site means (ASMs) for pairings of each of the three measures. The right columns show correlations of the across-site variances (ASVs) for pairings of each of the three measures. The three rows show correlations between MDTs in quiet and masked MDTs, MDTs in quiet and the amount of masking, and masked MDTs and the amount of masking, respectively. Symbols stand for different subjects, with the filled ones representing right ears, and the open ones representing left ears. Regression lines represent linear fits to the data. Correlation coefficients (r) and p values are shown in each panel.

Figure 4

Phoneme recognition using four MAPs as a function of SNR. Group mean (left) vowel and (right) consonant recognition using the subjects’ everyday-use clinical MAP as well as three experimental MAPs are shown as a function of SNR. Performance for different MAPs is shown in different symbols. Error bars represent ±1 standard deviation.

Figure 5

Individual and group mean (top) vowel recognition and (bottom) consonant recognition at 0 dB SNR. Error bars represent ±1 standard deviation. On the x axis, the MAP letters are shown for each subject, with the letter E representing the subject’s everyday-use clinical MAP. Asterisk indicates MAP B produced statistically significantly better group mean performance compared to the performance using the subjects’ everyday-use clinical MAP.

Figure 6

Individual and group mean speech reception thresholds (SRTs). SRTs represent the signal to noise ratio required for a listener to repeat 50% of the sentences accurately. Lower scores indicate better ability to recognize sentences in noise. For subjects who benefited from MAP A, an additional control MAP (Odd/Even) was tested and results are shown in the fifth (open) bar. For each subject, the best MAP is indicated by a labeled arrow. In the group data, the asterisk indicates statistically significantly lower SRTs using MAP B compared to the SRTs using the subjects’ clinical MAP.