Read about Starkey research studies in these interesting summaries of data and clinical observations.
Acceptable Hearing Aid Throughput Delay for Listeners
with Hearing Loss Under Noisy Conditions
Hearing aid processing, remote microphone communication, and the overall number of frequency bands each contribute to the throughput delay of digital hearing aids. We have previously reported that most normal-hearing participants found delays of at least 20 milliseconds to be acceptable in real-world, noisy listening conditions. The current study extends to our previous research to include participants with hearing loss.
Validation of a Hearing Aid Program Designed for Music Listening
A research study was completed to evaluate a new music program, consisting of a distinct compression architecture and fitting formula, designed to improve sound quality while listening to music.
Speech Intelligibility Differences between Bilateral
and Monaural Telephone Listening Conditions
Hearing aid users often report difficulty hearing on the telephone. The purpose of this study was to determine if bilateral listening via wireless ear-to-ear audio transmission is advantageous as compared to monaural listening for telephone conversations and to validate wireless ear-to-ear audio transmission as a viable option for hearing aid use for the telephone.
Maximum Acceptable Delay in Hearing Aids Under Noisy Conditions
All digital sound processing requires a brief delay. For hearing aids, the range of tolerable delay may vary with listening condition. This study assessed listener preference for noisy conditions, which were hypothesized to extend the range of tolerable delay.
Speech Perception and Electrophysiological Outcomes
with ReadMyQuips Auditory Training Program
This study assessed the role of auditory training in improving selective attention ability. Following regimented auditory training, participants showed some objective improvements in the averaged electrophysiological responses.
Is repeating audio helpful in real-world environments?
Experimental hardware was prepared that allowed for real-time audio recording through hearing aid microphones. These recordings could be played back by the listener through a mobile application. Participants opinion of this utility varied greatly with intent, acoustic event, and software implementation.
Side change harmonic enhancement of noise corrupted speech
for hearing impaired listeners
A novel technique for speech enhancement applied non-linear distortion to speech components isolated from a noisy mixture; these components were then mixed with the original signal. Participant judgments indicated an improvement in speech clarity with some sacrifice in overall sound quality.
Examining relationships between cognitive status
and demographic and audiologic factors
An analysis was completed to understand relationships between cognitive status and a series of audiologic factors in a sample of 115 participants with hearing loss. Regression models of varying complexity found mild relationships between the individual factors. A grand model, including all factors, accounted for 28% of the variability in screened cognitive status.
Examining relationships between cognitive status and hearing aid factors
An analysis was completed to understand relationships between cognitive status and hearing aid outcomes in a sample of 115 participants with hearing loss. Regression models of varying complexity found mild relationships between the individual factors. A grand model, considering all factors, accounted for 51% of the variability in screened cognitive status.
Validation of a new online inventory for assessing spatial hearing abilities
An online inventory of spatial hearing ability (ISHA) was developed and validated with a group of 78 participants. The questionnaire includes 27 questions that use a combination of contextual images and text. Analysis of sensitivity at the test and question levels suggest that the ISHA is a valid tool for the measurement of spatial hearing ability.
Preferred Aided Listening Levels for Music in the Sound Field
For six different hearing aids, participants adjusted music presented in the sound field to their preferred listening level. Results show high inter-individual variability in preferred listening levels but comparatively low intra-individual variability across hearing aids.
Cognitive and Hierarchical Models for Preference Data
A hierarchical Bayesian framework is used to aggregate and analyze partial preference data using a Thurstonian model. This approach overcomes the limitations faced by parametric assessment, allowing for the ordinal nature of the data and incompleteness in the data set.
A Hierarchical Model for the Analysis of Individual’s Data
Data collected in audiology research are often sparse in number of observations highly variable. Leveraging prior knowledge of performance, a hierarchical model is proposed for the generation of confidence intervals around an individual’s performance. Subsequently, statistical differences can be assessed between individual’s performance.
Relating Hearing Aid Gain Settings to Clinical Outcome Measures
Analysis of 68 hearing aid fittings showed that patients prefer gains that are lesser than those prescribed by the NAL-NL1 and similar to those prescribed by the e-STAT formula. Outcome measured show a moderate relationship to the amount of high-frequency gain prescribed.
Benefit of linear and non-linear amplification with time-varying maskers
A comparative comparison of linear and WDRC hearing aid fitting suggest that linear fitting may result in lower listening effort, while WDRC fitting improves speech recognition ability.
Incorporating Spatial Information into Binaural Noise Reduction
A method for binaural noise reduction was assessed that incorporates prior knowledge of spatial information. The system proved to be more accurate when compared to a monaural implementation.
Changes in Event-Related Potentials after Completing the ReadMyQuips Auditory Training Program
New hearing aid wearers completed auditory training following the first hearing aid fitting. Some individuals showed stronger event-related potentials following hearing aid use and auditory training.
Selective Attention in Speech Understanding with Competing Talkers
A task of selective attention was used to assess contributing factors in tracking sound recognition and identification of meaning in a sentence. Audibility was found to account for recognition of sounds while linguistic and cognitive abilities appear to account for ones ability to identify meaning within a sentence.
Spatial Release of Cognitive Load
Cognitive requirements of listening in multi-talker environments were assessed. Increasing spatial separation between talkers reduces listening effort. A conceptual model for cognitive demand while recognizing speech among multiple talkers is proposed.
Improving Speech Understanding in Adverse Listening Environments Using Distributed Microphones
An algorithm for distributed microphone processing is proposed. Simulation results show improvements in speech understanding that surpass traditional approaches.
Real-World Recording Database for Distributed Microphone Research
Cocktail party recordings were completed for multiple target talkers in a background of conversations that included 8, 24, and 56 competing talkers. Analysis of the recordings confirms the integrity of this corpus for future with with complex microphone arrays.
Viability of RECD in fitting vented and open-canal hearing aids
Hearing aid fittings were completed using real-ear to coupler differences (RECD) measured in different coupling configurations. These configurations ranged from fully occluding to an open-canal. As the hearing aid coupling configuration became more open, the accuracy of the RECD fitting method decreased.
Do combination instruments reduce the effects of tinnitus?
Participants with significant tinnitus handicap were prescribed traditional hearing aids and hearing aids with a sound therapy feature. Significant benefits of hearing aid use and sound therapy were observed in outcome measurement.
Speech recognition in noise with four remote microphone technologies
Four different wireless remote microphone technologies were used during a task of speech recognition in noise. All of the tested remote microphones significantly improved speech recognition in noise.
Acoustic variability of occluded earbuds in receiver-in-the-canal hearing aid fittings
The occluding characteristics of different instant-fit silicon earbuds was assessed through real-ear measurement. High-variability in measured occlusion was observed between participants. This suggests that any single type of earbud may be significantly more occluding for some hearing aid wearers than others.
Effects of aging and spectral shaping on brainstem differentiation of consonants
Aging reduces responsiveness of the brainstem to dynamic cues of speech. This study found that enhancement of time-varying amplitude cues, around the second formant frequencies, can reduce this age-related deficit. It is speculated that such enhancement may free sub-cortical resources for application to ancillary tasks.
Assessing the contribution of spectral cues to recognition of frequency lowered consonants
Frequency lowered speech cues were spectrally enhanced in an effort to reduce consonant confusions. A proportion of the participant sample showed significantly reduced confusions with preserved and enhanced spectral cues.
Derivative free optimization of hearing aid parameters
A modified model for estimation of loudness perception is proposed for use in prescription of target hearing aid gains. This model accounts for contribution of sound at distant frequencies in the overall prediction of loudness. The result is a prescriptive model for hearing aid gains that accounts for spread of excitation while restoring specific loudness.
Annoyance perception for hearing impaired listeners: A revisit
Acoustic properties of sound that dominate the perception of annoyance in listeners with hearing loss were identified. Loud and narrow band, high-frequency sounds are the most likely to be judged as annoying. These findings will assist in the development of strategies for classification and management of annoying sounds with hearing aids.
Acceptable Noise Level: Effect of Presentation Level, Digital Noise Reduction, and Stimulus Type
In this collaboration with the University of South Florida significant improvements in the Acceptable Noise Level were observed when listening with digital noise reduction (Voice iQ2) in a background of speech babble. Further, the data suggest that participants with greater sensitivity to noise experience the greatest benefits from digital noise reduction.
Clinical Validation of Multiflex Tinnitus Technology
Sound therapy, the use of sound for the purpose of tinnitus management, is widely accepted as a management tool for tinnitus. This poster presentation reports on the clinical validation of MultiFlex Tinnitus Technology sound therapy.
Improve Resistance to Foreign Material for Hearing Instruments
Nano-coatings have greatly improved the moisture resistance of hearing aids. This poster describes the hydrophobic properties of nano-coating and reviews results from an independent laboratory that tested a series of hearing aids for ingress of water and cerumen.
Optimization of Loudness Restoring Hearing Aid Fittings
A number of hearing aid fitting algorithms prescribe gain based on estimates of loudness perception. A modified prescription that better accounts for spread of excitation and inner or outer haircell loss is proposed.
Perceptual Evaluation of a Binaural Beamforming Algorithm
Traditional binaural beamforming degrades localization cues. This study demonstrated benefits for a binaural beamformer that retains localization cues while maintaining benefit over unlinked, unilateral beamformers.
Large-Scale Objective Assessment of Preferred Gain and Frequency Response for Multiband Compression Hearing Aids
In a study of 127,000 hearing aid fittings, it was observed that 50% of final hearing aidfittings fall within +/- 4dB of the selected prescriptive target. Fittings for severe hearing loss showed greater variability than those for milder hearing loss.
Characterization of Multi-Class Acoustic Environments
Real-world environments consist of multiple sound sources that make classification of single sounds (e.g. speech or noise or music) inefficient in hearing aid signal processing. This study developed an idealized method for classifying acoustic environments with multiple sound sources. Crowd-sourced methods for data collection were leveraged in the development process, validating the utility of Amazon’s Mechanical Turk in hearing science applications.
Comparison of Adaptive Versions of the CCT and NST
Adaptive speech recognition tasks vary speech or noise levels in response to correct or incorrect responses, with a goal of maintaining 50% performance. This studyshowed efficacy in implementation of the CCT and NST (two well established speech recognition tests) in an adaptive test format.
Design and Evaluation of a Bio-Inspired Compression System
A compressor that mimicked loudness growth characteristics of the normally functioning cochlea was evaluated. Specifically, channel bandwidth was narrower at low input levels and wider at high input levels. This level-dependent channel behavior is hypothesized to restore loudness growth to near-normal rates while retaining high-frequency spectral contrasts that may be negatively affected by traditional compression architecture.
On Factors Affecting Speech Reception in a Multi-Talker Listening Scenario
Using a complex, spatially distributed, speech-in-noise test, factors to determine what may be a valid predictor of a person’s ability to understand speech in challenging environments were investigated. It was determined that processing speed and executive function are important non-audibility factors in the understanding of speech.
Applications of an Annoyance Perception Model to Noise Reduction for Hearing Aids
Annoyance perception for listeners with hearing loss differs from that of normal hearing listeners. A mathematical model to predict annoyance perception was developed and applied to the development of a digital noise reduction algorithm. The digital noise reduction that targeted annoyance reduction provided clear differences when compared to traditional, energy-based digital noise reduction.
Annoyance Perception and Modeling for Hearing Impaired Listeners
Perceptual annoyance of environmental sounds is measured for normal-hearing and hearing-impaired listeners under iso-level and iso-loudness conditions. Data from the hearing-impaired listeners shows similar trends to that from normal-hearing listeners, but with greater variability across individuals. A regression model based on the statistics of specific loudness and other perceptual features is fit to the data from the normal-hearing listeners, and is used to predict annoyance for the hearing-impaired listeners. Differences across the subject populations are discussed.
Clinical Factors Affecting Directional Benefit
Benefit with directional microphones can be variable across individuals with hearing impairment. It is assumed that directional benefit can be predicted by electroacoustic measures such as the directivity index but the relationship between directional benefit and other clinical factors is less clear. This poster presentation details an analysis of directional benefit for individuals fit in either open or occluded configurations as measured by the Hearing in Noise Test as a function of various clinical factors.
Detection of word final /s/ and /z/ with three modern frequency lowering technologies
In the prescription of hearing aids, providing insufficient high-frequency gain may have a negative impact on detection of phonemes such as word final /s/ and /z/. Modern techniques for frequency lowering offer an opportunity for improving detection of these phonemes. In recent years three techniques for frequency lowering have been introduced to commercially available hearing aids. This poster details a study that examined behavioral outcomes with three commercially available frequency lowering algorithms in a group of adults with sloping severe high-frequency hearing loss.
Openness of fit and benefit from adaptive features
Openness of the coupling between the hearing aid and ear modifies the acoustic properties of a hearing aid fitting. An open-canal hearing aid offers benefits related to comfort and reduction of perceived occlusion for the patient. The open-canal hearing aid also limits attainable low-frequency levels in the ear. There is also a greater contribution of direct unamplified sound that enters the ear when compared to more occluding hearing aid fittings. This poster details a study that documented comparative benefits of digital noise reduction and directional microphones with increasing openness of fit.
Evaluation of objective and subjective benefit with a novel frequency lowering algorithm
This research poster details a study in which subjects with varying degrees of hearing loss were fit with custom or receiver-in-canal (RIC) hearing aids and the frequency lowering algorithm (Spectral iQ; Starkey Hearing Technologies). Spectral iQ uses real-time spectral envelope warping to reproduce high-frequency spectral information in a lower frequency region where audibility can be achieved. Over a six-week clinical evaluation, objective performance and subjective outcomes were measured with a variety of laboratory and field tests, including a word-final consonant identification speech test and subjective outcome measures evaluating perceived sound quality and speech understanding.
Will Digital Noise Reduction Improve Successful Hearing Aid Use?
Predicting hearing aid users' success with amplification is difficult. A metric of a listener's tolerance to background noise, known as the Acceptable Noise Level (ANL) test has shown promise in predicting success with hearing aids; additionally, noise reduction technology has been effective in improving listeners' acceptance noise. If hearing aid success can be predicted with the ANL test and a relationship exists between ANL scores and noise reduction, clinicians may be able to select optimal noise reduction settings that will maximize a patient's success. This study investigated this relationship between digital noise reduction and outcomes with the ANL.
Can Acceptable Noise Levels Be Predicted from a Noise-Tolerance Questionnaire?
Historically it has been difficult to predict success with hearing aids. Recent research has offered hope that hearing aid success may be predicted with a test of the Acceptable Noise Level (ANL); a measurement of the amount of background noise that a listener is willing to accept. A questionnaire was developed to gain insight into the cues that listeners are using to determine their individual ANLs. Understanding these cues may offer an explanation as to why some individuals are successful with hearing aids while others are not.
Can Directional Benefit be Predicted from Auditory Threshold?
Forty-four research participants completed a study that evaluated speech understanding in noise while wearing S Series™ hearing aids. One of the experimental questions asked if directional benefit in a noisy background — relative to performance with omni-directional microphones — could be predicted from the pure-tone audiogram. While trends within the group data suggest that higher pure-tone thresholds may result in the greatest benefit from directional microphone technology, individual variability in performance suggests that directional benefit should not be attributed to audiometric data alone.
Real-Ear Measurements Improve Subjective Performance of Hearing Aid Fittings
This article documents the real-world benefits of incorporating real-ear measurements into clinical practice. Data collected from clinical sites using real-ear measurements were compared to data from sites that did not complete real-ear measurements. Fittings optimized using real-ear verification resulted in increased high‑frequency output relative to fittings that did not utilize real-ear verification. Data from the Device‑Oriented Subjective Outcome (DOSO) Scale indicate that the fittings optimized with real-ear measurements delivered significantly greater benefit than fittings that were not optimized through real-ear measurements.