Optimizing hearing aid algorithms to improve spatial SRM
One of the key concepts in hearing and hearing aid development is spatial release from masking, or SRM. When listening to a speaking person in an environment with background noise, the target sound signal is partially ‘masked’ by noise. The human brain uses the competing sources’ direction and frequency spectrum to differentiate between them and understand speech. The listener experiences spatial release from masking: due to the awareness of the spatial distribution of the sound sources, the target signal is more easily distinguished from the background noise. The result: the listener perceives a higher signal to noise ratio (SNR).
A hearing disorder can throw a spanner in the works and impact speech intelligibility however. Especially older adults and adults with sensorineural hearing loss (SNHL) may have an impaired capacity for SRM. In general, hearing aids perform poorly in mitigating this.
Marc Brennan & Ava Feller from the Special Education and Communicative Disorders participated in the ASA 2021 Tympan Challenge. Their submission focused on exploring how hearing aid algorithms can be optimized to improve spatial SRM. They used the Tympan to evaluate the effect of the number of compression channels on SRM. Tympan Rev-E (with a fast 600 MHz Teensy processor) was used for their experiments, a suitable platform for real time processing of the required multi-channel algorithms. Nine test subjects were presented with a target signal (AzBio sentences) and background noise in a test setup that allowed for spatial and spectral variation (see image).
The reasoning for the experiments is as follows: A single channel algorithm processes the incoming sound in the complete frequency range the same. Using compression and expansion algorithms, the incoming sound signal is treated to emphasize (a.o.) the interaural differences used to establish the sound's source and direction, or spatial cues. A multichannel algorithm splits the signal into frequency bands, and treats each band separately. The hypothesis for the experiment is that the multichannel treatment helps to improve these spatial cues, making them more audible. More channels should provide a higher spectral resolution, thus a more detailed improvement of the spatial cues
Of course, this principle will work better for some situations than others, for instance the character of the background noise. Increasing the number of compression channels works differently in a situation where the masking sound has, for instance, high information content (speech-in-speech masking) in comparison to low information content (like the sound of a machine: speech in noise masking or energetic masking). Also the improvement by processing the sound was expected to occur predominantly with the participants with a hearing impairment.
The results give some insights in the effects of amplification, level of the masker signal and the amount of compression channels on spatial release from masking. Furthermore, the study provides a proof of concept for Tympan as a suitable research platform for these types of studies.
About the TYMPAN ASA2021 Design Challenge
During the ASA (Acoustical Society of America) conference in June 2021, Tympan hosted a design challenge: What is possible with the Tympan?
10 exciting new applications were submitted and presented at the following ASA conference: Enhancements of hearing aids, spatial acoustic processing and smart earphones and much more. Stay tuned if you want to learn what is possible and to keep track of future developments.