Cochlear implants (CIs) take advantage of the tonotopic structure of the cochlea, stimulating areas progressively closer to the base as the input sound frequency increases. One key fitting parameter in a CI is the frequency table wherein input acoustic frequencies are allocated to intracochlear electrodes. Under current clinical practice, all users of a given CI model receive basically the same frequency table. In postlingually hearing impaired CI users, this one-size-fits-all approach may introduce mismatches between input acoustic frequency and the characteristic frequency of the neurons that are stimulated. Although human listeners can adapt to these distortions, there is growing evidence that sometimes this adaptation process may be incomplete. Both the adaptation process and its possible incompleteness have important consequences to speech perception in the postlingually hearing impaired CI population.
The first aim of the proposed work is to measure the extent and the time course of this adaptation process, which will be explored in two experiments. Experiment 1 will study recently implanted CI users and follow them for a year, using four different methods to measure adaptation to frequency mismatch, complemented by a battery of speech perception, psychophysical, cognitive, and anatomical measurements. Experiment two will examine the same questions but with a group of CI users with long term experience (at least one year). Both experiments will also have a second aim: using the anatomical and cognitive measures to predict which individuals are more likely to suffer incomplete adaptation to their clinical frequency tables. Lastly, Experiment 3 will address the third aim of the proposed research: to test the hypothesis that frequency tables intended to reduce frequency mismatch will improve speech perception scores in those CI users who show incomplete adaptation. Part of the proposed work involves developing and refining software and hardware tools to facilitate the search for alternative frequency tables that may help minimize frequency mismatch. In summary, the experiments described in this proposal will provide new insights about the nature of auditory adaptation to a modified peripheral frequency map by postlingually hearing impaired CI users, and will also provide guidance to the clinicians who are in charge of fitting these devices. Studies like the present ones will help translate basic knowledge into clinical practice, and will make clinical practice more data- and theory-driven.
The proposed research aims to understand how human listeners adapt to distortions in their auditory frequency maps. This knowledge may help improve speech perception in cochlear implant users who have difficulty adapting to the standard frequency tables programmed into their speech processors. The project also aims to develop the hardware and software tools that will allow the translation of these scientific findings into actual clinical practice. Project Narrative The proposed research aims to understand how human listeners adapt to distortions in their auditory frequency maps. This knowledge may help improve speech perception in cochlear implant users who have difficulty adapting to the standard frequency tables programmed into their speech processors. The project also aims to develop the hardware and software tools that will allow the translation of these scientific findings into actual clinical practice. __SpecificAimsTextDelimiter__ A. SPECIFIC AIMS Cochlear implants (CIs) use frequency tables that determine the allocation of acoustic frequencies to different electrodes in the cochlea. These frequency tables mimic the tonotopic structure of the cochlea by stimulating areas progressively closer to the base as the input sound frequency increases. However, the standard frequency tables normally used in clinical practice do not account for factors such as electrode location, cochlear size, or neural survival. As a result, in postlingually hearing impaired (HI) CI patients there is no guarantee that an input signal with a given acoustic frequency will stimulate neurons tuned to that same characteristic frequency. Instead, distortions such as frequency shift, compression, or expansion may be introduced. Current clinical practice is based on the assumption that postlingually HI CI users successfully adapt to these frequency distortions, despite individual differences in cochlear size, neural survival, electrode location, and residual acoustic hearing. If this assumption was true, the one-size-fits-all approach would make sense. However, what little information we have available suggests that adaptation to frequency distortions, though substantial in many CI users, may not always be complete. Both the adaptation process and its possible lack of completeness may have important consequences for speech perception in the postlingually HI CI population. Initial frequency mismatch may affect speech perception scores shortly after initial stimulation. As listeners adapt to the mismatch, speech perception scores improve. However, CI users who do not adapt completely to this frequency mismatch even after prolonged experience may reach an asymptotic speech perception level that is suboptimal (in the sense that it would be higher after complete adaptation). The proposed work involves four complementary methods to measure whether adaptation to the frequency table used by a given CI user is complete. The first method will involve measuring listeners' 'perceptual vowel space' (Harnsberger et al., 2001; Svirsky et al., 2004). With this task, listeners are required to select synthetic vowels that best match a given target vowel. The vowels are selected from a 2-dimensional grid where they are arranged in terms of increasing F1 and F2 as one moves from the lower left to the upper right corner. When using this method, lack of adaptation to a frequency table is demonstrated by deviations from the norm with regard to the location and the size of the regions selected to match each vowel. In the second method we will use a mathematical model of vowel identification (Sagi et al., 2009; Sagi et al., in press) to determine the location and size of listeners' phonetic labels (or 'response centers') for vowels. As with the first method, lack of adaptation to a frequency table is revealed by the size and location of the response centers. The third method will involve pitch matching measures between electrical and acoustic stimuli in a subgroup of subjects, those who have usable residual hearing. In this case mismatch is indexed by the difference between the frequency band assigned to a given electrode and the acoustic frequency that is perceived to match the pitch elicited by that electrode. The fourth method will involve obtaining 'listener- selected' frequency maps obtained with a special tool that allows CI users to change map parameters (mean frequency and total bandwidth of the filter bank) in real time in order to maximize speech intelligibility. When CI users select a frequency map that differs from their clinical frequency map, we infer that adaptation to the clinical frequency map is incomplete. To help interpret the adaptation data obtained with these four methods, we will also obtain measures of speech perception, formant frequency discrimination, cochlear size, electrode location, verbal learning, working memory, and subjective judgments. These four methods for measuring adaptation to frequency maps, when coupled with our additional measures, will allow us to address three specific aims. Specific Aim 1 is to measure the extent and time course of adaptation to frequency-to-electrode tables in postlingually hearing impaired CI users. This aim will be addressed with two experiments, a longitudinal one with recently-stimulated CI users, and a cross-sectional one with experienced CI users. Specific Aim 2 is to test the hypothesis that incomplete adaptation to a frequency table (measured with each one of the four methods listed above) is more likely in cases of large cochleas, shallow electrode insertion, low verbal learning skills, low levels of working memory and may be affected by the presence of usable residual hearing. Specific aim 3 is to test the hypothesis that frequency tables intended to reduce frequency mismatch will improve speech perception scores in those CI users who show incomplete adaptation to their standard clinical frequency tables. The proposed experiments will investigate basic aspects of adaptation to different frequency tables after cochlear implantation in postlingually HI listeners. These experiments will also have an important translational aspect, as they will try to predict (based on anatomical, cognitive, and psychophysical measures) which listeners may have most difficulty adapting to frequency mismatch. Even more importantly from a translational perspective, we will investigate a possible way to mitigate the effect of such frequency mismatch. In so doing, the present studies will provide important basic knowledge about perceptual learning as well as useful and specific guidance to the clinicians who are in charge of fitting CIs.
Showing the most recent 10 out of 47 publications
