Evaluation of a New Noise Reduction Strategy in Hearing Aids

Test measures speech perception in the presence of 5 competing talkers, presented from 5 speakers located in front of the participant. The talkers are presented sequentially with a brief offset of 1.2 seconds. One of the talkers is the target talker and the participant is instructed to listen for a specific call sign, which identifies the target talker. The participant is prompted to identify the color and number that is subsequently spoken by the target talker. The location of the target talker is randomized from the five locations in the front hemisphere. A score of correctly repeated color and number is calculated and a higher score is better. The dB levels at which the target talker and competing talkers are presented are determined individually for each participant by the Speech in Noise Test, described in Outcome 4.

An adaptive method of measuring subjective listening effort. Three sentences are presented sequentially in noise. The listener rates their subjective effort on a 13-point scale that goes from Effortless (1) to Extremely Effortful (13). There is a 14th point which is labeled "Only Noise" and indicates no speech was heard. After each rating, the volume of the speech adapts. The algorithm is designed to capture the SNR (Signal to Noise Ratio) that includes both the effortless and the extremely effortful rating for an individual. Listeners rate different SNR levels using the scale, for instance, a listener may rate an SNR of -12 dB (where speech is 12 dB softer than the noise) as Very Effortful, or 11 on the scale, while rating a SNR of -2 (where speech is only 2 dB softer than noise) as Effortless or a 1 on the scale. In comparing HA settings, that which has the least negative SNR for a particular rating on the scale indicates less effort.

Participants ranked the three devices on a scale from 1 (worst) to 5 (best) for five characteristics: Noise suppression, overall satisfaction, speech clarity, speech loudness, speech naturalness. For each participant, ranks were assigned to each device based on the rating it received, with 1 being the highest rank and 3 being the lowest rank. The frequency of each rank was then calculated for each device across all raters. Results are displayed in the count of devices in each ranking. No statistical analysis was performed on this measure.