Effects of Nonlinear Signal Processing Algorithms on Speech Perception

The purpose of this study is to investigate the effect of nonlinear signal processing algorithms on speech perception.

In this study, we are interested in how nonlinear algorithms influence speech perception. Nonlinear algorithms are used within hearing aids and personal sound amplifiers to provide comfort to the listener. One of these algorithms provides amplification (gain) in a manner that depends on the level of the input sound.: Low-level sounds are amplified much more than high-level sounds. This type of amplification makes sounds more comfortable for listeners, but also distorts incoming sounds. To determine the effect of these algorithms on speech understanding, we will evaluate the following factors on speech perception, within the context of these algorithms: The input Signal-to-Noise ratio (SNR) to the algorithm (3 different SNRs) The type of background noise (20 people talking or 2 people talking) Algorithm settings (slow and fast - whether changes to gain applied to fluctuating input sounds occurs quickly or slowly) To address our questions, we are using a wearable processor (Tympan) that allows for real-time processing of audio signals. The Tympan itself contains microphones, a processor, and small earpieces that include speakers. A listener can wear the earpieces, like headphones, and listen to sounds processed by the Tympan in real time. The Tympan allows us access to the algorithms and to the sounds processed by the algorithms so that we cannot algorithm behavior to speech perception. Our study will complement existing work on speech perception and nonlinear algorithms, but our study will be the first to use a wearable processor in which the specific algorithm is known, tailored to an individual's hearing levels, and which allows for direct calculation of the output SNR. Most studies also have not measured speech perception, and this will be one of only a handful of studies with that objective.

Perception will be measured for different algorithm settings and environmental variables (type of noise and signal-to-noise ratio)

Inclusion Criteria: Between the ages of 18 and 65 Native speaker of English Have bilateral, symmetric sensorineural hearing loss that is less than 75 decibels (dB) Hearing Level (HL) at 1000 Hz and below. Exclusion Criteria: Subjects with normal hearing, mixed hearing loss, or asymmetric sensorineural hearing loss. Subjects who younger than 18 or older than 65. Subjects who are not native speakers of English. Subjects with thresholds more than 70 dB HL at 2000 Hz and below