RP302 - Exploring Impact of Sensor-Based Directionality on Real-World Listening in Children

Abstract:
Adolescents with hearing loss listen in academic and social situations with multiple talkers and noise sources. We examined sensor-based directionality that incorporates acoustic and mechanical data from the hearing aid to modify the directional response of the hearing aid based on head and body movement. Sentence recognition was measured with and without sensor-based directionality for eighteen adolescents with hearing loss (12-18 years-old) in conditions that differed by target and masker location. Participants also completed a two-week trial with ecological momentary assessment with and without the feature. Benefits for sensor-based directionality were apparent in speech recognition in noise.

Summary:
Adolescents with hearing loss who use hearing aids are increasingly in listening situations that require them to attend to a specific talker when there are other interfering talkers. Although hearing aid features like directional microphones are designed to help improve speech recognition in these listening situations, traditional directional microphone systems adapt based on the acoustic characteristics of the signal from the hearing aid microphone(s). More recently, mechanical sensors that can contribute information about body and head movement were developed to improve the adaptability of directionality. However, sensor-based directionality has not been assessed in adolescents with hearing loss in laboratory or real-world settings. In this study, we examined laboratory speech recognition and real-world ecological momentary assessments (EMA) to assess the efficacy of sensor-based directionality for adolescents with hearing loss.

Eighteen adolescents with mild to severe hearing loss were fitted with bilateral hearing aids verified to DSL-5 targets. Participants and their parents were blinded to whether the sensor-based directionality feature was activated. The order of feature activation was randomized in both laboratory and EMA phases. In the laboratory phase, masked speech recognition was assessed with and without the hearing aid feature active in three spatial conditions: 1) Fixed co-located: target and masker stimuli presented from 0° azimuth; 2) Fixed separated: target presented from 0° azimuth and masker from +/– 54° azimuth; and 3) Random: target stimulus presented randomly from one of five locations and masker fixed at +/- 54° azimuth. Following laboratory speech recognition, the participants completed a two-week trial with the devices, one week with and one week without the sensor-based directionality. Android phones prompted the participants to complete 8 EMA surveys per day that asked about the listening environment, hearing aid use, tiredness, and listening difficulty. The sensor-based directionality feature was activated/deactivated midway through the trial using remote programming.

Linear mixed effects models were used to assess the effect of feature and condition on the speech recognition threshold for 50% correct (SRT50) controlling for age and aided audibility. There was a main effect of age, where the SRT50 decreased as age increased across conditions. The main effect of aided audibility on SRT50 was not significant. The interaction between condition and feature was significant. Post-hoc analyses indicated that the SRT50 for the fixed separated and random separated conditions were lower than the fixed co-located condition, consistent with spatial release from masking when target and masker are separated. The feature on and feature off conditions were only significantly different in the fixed separated condition with a 1.5 dB improvement in SRT50 when the feature was activated. The EMA data showed wide variability among participants with feature activation.

Adolescents with hearing loss can benefit from sensor-based directionality for speech recognition in noise in listening conditions where the talker and the masker are spatially separated, but the magnitude of benefit may depend on the spatial configuration of the target and masker stimuli. EMA can provide supplemental information about the real-world impact of blinded feature comparisons for adolescents with hearing loss.

Brief Summary of Clinical Takeaways:
The clinical takeaway for this session is that sensor-based directionality can increase speech recognition in difficult listening environments for adolescents with hearing loss. Additionally, use of EMA surveys can provide additional insight to everyday listening experiences for this population.