Study of Deep Neural Network Noise Reduction in Hearing Aids for Adults with Mild to Moderate Cochlear Hearing Loss Testing Speech Understanding in Different Background Noise Levels

What this Trial Is About

This research aims to understand how different noise-reduction technologies in hearing aids affect a persons ability to hear speech in noisy environments. The study focuses on people with mild-to-moderate cochlear hearing loss and compares various noise-reduction algorithms, including stronger and weaker settings, as well as no noise reduction, across different background noise levels. The goal is to identify which noise-reduction strategies help improve speech understanding the most, especially in challenging listening situations. Participants will complete listening tasks using a dual-sentence method that simulates real-world listening demands by requiring them to recall two sentences presented back-to-back amid background noise. Each participant experiences all noise-reduction settings off, low, high at different signal-to-noise ratios SNRs, ranging from easier conditions where speech is louder than noise to harder ones where noise dominates. The study uses a monaural, single-microphone hearing aid configuration to isolate noise-reduction effects without directional hearing aid processing. During the study visit, which lasts about 1 to 2 hours, participants speech understanding is assessed by scoring how accurately they recall keywords from the sentences under each condition. This allows researchers to measure both immediate and delayed speech recall as indicators of speech intelligibility and cognitive effort. The findings may help improve hearing aid design and guide clinical fitting by showing how different noise-reduction algorithms impact speech comprehension and listening difficulty in everyday noisy settings.

Evaluating a Deep Neural Noise-Reduction Algorithm for Hearing Aids