Using psychometric procedures for estimating Minimum Audible Angle thresholds, the study demonstrated that increased spectral complexity and higher presentation levels significantly enhance spatial resolution in normal-hearing individuals.
Audio Language Models are highly capable classifiers of hearing difficulty moments in conversational audio.
Investigating Individual Aspects of İnnovativeness and Attitudes Towards Artificial Intelligence in Audiology Students and Graduates
The current study employs stimuli that are reconstructed by vocoding stimulation patterns generated by an actual CI, then presented to NH listeners to determine psychometric curves for VTL difference.
A computational model that simulates loudness perception experiments using a peripheral auditory model for electric and acoustic stimulation (PAMEAS).
This study evaluates the accuracy, reliability, and readability of AI chatbot responses across five key domains of hearing health, highlighting their potential and limitations in patient education.
Validation of bone conduction hearing threshold estimation at 500, 1000, 2000, and 4000 Hz using Rapid Cortical Auditory Evoked Potentials from 12.5-minute water-based EEG recordings.
A generative AI platform that lets clinicians interact with lifelike virtual patients to build communication skills through realistic, real-time conversations.
This study investigates neuroanatomical changes in a military population, aiming to enhance our understanding of how tinnitus and hearing loss affect the brain's structure.
This study aimed to assess the validity and feasibility of patients using smartphone hearing test applications to self-assess hearing thresholds at home, with comparison to pure tone audiometry.
This study introduces a new method for designing optimized multipolar stimulation strategies, offering improved auditory perception with potentially reduced energy consumption.
Does applying dynamic range compression separately to speech and noise improve intelligibility compared to joint compression, when using separated sources in non-stationary noise?
Incorporating Wikipedia-derived excerpts influenced the readability of AI-generated hearing health responses, with more pronounced effects in Portuguese, where texts remained consistently more complex across metrics than English content.
This study introduces AHEAD-DS, a new sound dataset designed for scene recognition of audiologically relevant environments, and YAMNet+, a sound recognition model.
By highlighting the correlation between otitis media rates and social determinants such as housing and healthcare access, this mapping tool aims to provide targeted community level information for policy and service delivery in remote communities.
A lightweight adversarially trained end-to-end model improves speech clarity for cochlear implant users in noisy environments without increasing system complexity.
This study evaluated the importance and effectiveness of using digital technology and reusable materials to provide teleguidance and telemonitoring to new and experienced hearing aid users and its impact on the performance and daily use of hearing aids. The individuals were followed for a period of eight to 12 months. The users' performance was monitored using the Speech Spatial and Qualities of Hearing Scale questionnaire.
This work examines how neural signals recorded during infancy can serve as early biomarkers for predicting speech, language and learning difficulties in neurodiverse populations.
This study compares cochlear implant processing strategies and finds that the HSSE method preserves both spectral and envelope cues in a balanced way, making it most similar to natural sound.
In our study, the effects of different presentation methods (live sound, recorded sound, etc.) of stimuli on performance in free-field speech tests performed on individuals using bilateral hearing aids were examined, and significant data were observed in some groups among the presentation methods.
Artificial Neural Networks generate human-like speech perception and inform the effect of context on human speech recognition.
This project aims to assess the impact of reverberation on spatial acuity by measuring the minimum audible angle (MAA) in child and young adult bilateral cochlear implant users in simulated reverberant environments.
Tiresias is an open, low-power, and wearable-ready platform that enables the development and deployment of hearing aid algorithms using consumer hardware, visual programming, and open-source tools—bridging the gap between clinical needs and embedded system development.
Distortion product otoacoustic emissions (DPOAE) signal-to-noise ratios in adolescents offer a promising, underutilized tool for both detecting and predicting high-frequency hearing loss, paving the way for earlier intervention against noise-induced damage.
Using machine learning, an analysis of a large clinical audiology database, throws light on clients who have normal hearing thresholds, but report hearing difficulties.
Typical hearing participants show perceptual learning but not sensorimotor learning with real-time noise vocoded speech auditory feedback during speech production.
We develop methods that (1) enable accurate design of auditory filterbanks based on desired specifications for peak frequencies, bandwidths, and delays and (2) allow for very systematically studying the effect of varying filter characteristics on perceptual tasks and on improving speech and hearing technologies.
A novel direct-drive hearing device actuator (DHDA) was developed to enhance low-frequency compensation and sound fidelity while providing a non-invasive alternative for treating sensorineural hearing loss.