PP311 - An Additive Approach to Evaluating Aided Speech Perception in Noise

Abstract:
An additive approach was used to compare unaided and aided speech perception and subjective ratings with amplification alone, amplification plus digital noise reduction, and amplification plus digital noise reduction and adaptive directional microphones to determine how much each feature contributes to overall performance in noise. Results suggest that objective performance, as well as subjective ratings of listening effort, speech clarity, and sound quality, are largely driven by amplification and directionality, although the degree to which each algorithm contributes may depend on unaided speech performance in noise rather than the severity of the hearing loss.

Summary:
Patients with hearing loss commonly complain that even with hearing aids, they have difficulty hearing noise. From the MarkeTrak 2022 survey, respondents indicated lower overall satisfaction with the sound of their hearing aids than with the physical aspects. This was particularly seen in satisfaction with the hearing aids’ ability to minimize background noise (Picou, 2022). To combat background noise, most modern hearing aids have both directional microphones and digital noise reduction. Directional microphones are known to provide benefits for speech understanding in background noise (Ricketts & Hornsby, 2006; Walden et al., 2005). While digital noise reduction features alone rarely show a benefit for speech understanding, when partnered with directional microphones, the combined features may provide other benefits, such as decreased objective listening effort (Desjardins, 2015).
The purpose of this study was to assess the additive effects of amplification, digital noise reduction algorithms, and adaptive directional microphones, on both objective and subjective outcomes. Forty-three participants with an average mild sloping to moderately severe sensorineural hearing loss were fitted with receiver-in-canal or custom premium hearing aids with directional microphones. After wearing the devices in the field for approximately six weeks, speech recognition in noise and subjective ratings were measured under the following conditions: unaided, aided with noise reduction features turned off and microphones in omnidirectional mode (NR Off + Omni), aided with noise reduction features at the default settings and microphones in omnidirectional mode (NR On + Omni), and aided with noise reduction features at the default settings and microphones in the default adaptive-directional mode (NR on + Dir). Participants were seated in a sound booth in the center of an 8-speaker array with speech presented from the front and uncorrelated babble noise from the remaining seven speakers (summated to 67 dB A). Participants repeated 20 IEEE sentences per condition. Sentences were first presented in the NR Off + Omni condition to calculate an SNR50, or the presentation level at which 50% of the sentences were repeated back correctly. The mean presentation level across all sentences in this initial NR Off + Omni condition was then used to test each of the four experimental conditions in randomized order. After completing each test condition, participants were asked to rate perceived listening effort, sound quality, and speech clarity.
Results indicated a statistically significant difference in speech understanding scores (χ² (3)=49.99, p< 0.001) across conditions. Post-hoc analyses (Wilcoxon Signed-Rank Tests) revealed improvement in speech understanding in noise for all aided conditions (NR Off + Omni, NR On + Omni, and NR On + Dir) over unaided and an improvement in speech understanding in noise for NR On + Dir over the other aided conditions (NR Off + Omni and NR On + Omni). Results suggest that objective performance, as well as subjective ratings of listening effort, speech clarity, and sound quality, are largely driven by amplification and directionality, although the degree to which each algorithm contributes may depend on unaided speech performance in noise rather than the severity of a hearing loss.