Spatial release from masking in hearing impaired listeners
thesisposted on 28.03.2022, 01:15 authored by Baljeet Rana
The improvement in speech understanding in noise due to the spatial separation of a target source from masking sources is known as spatial release from masking (SRM). SRM observed when fluctuating distractors are present on both sides of the listener, and target speech at the front, can be largely attributed to a phenomena known as better-ear glimpsing (BEG). BEG, which utilizes interaural level differences (ILDs) to help understanding speech in noise, is limited by reduced audibility at high frequencies in hearing-impaired (HI) listeners. Unfortunately, it is difficult to provide adequate amplification at high frequencies to fully restore BEG. Therefore, the idea of extending ILDs to low frequencies was proposed here. The results of a first study showed that ILDs can be effectively utilized at low frequencies by both normal-hearing (NH) and HI listeners. However, the performance noted in HI listeners was still poorer than in NH listeners, which might have been due to differences in audibility between the two groups. To test this hypothesis, in a second study, both groups were tested at different equal audibility levels. Results showed that if audibility is carefully controlled then HI listeners can utilize ILDs as effectively in BEG as NH listeners. However, it was also observed that not all HI listeners could accommodate the required signal levels due to loudness discomfort. Therefore, in a third study, the effect of wide dynamic range compression (WDRC) on the utilization of ILDs was investigated. The results revealed that low-frequency extended ILDs combined with WDRC can provide BEG cues that are significantly higher than the one provided by natural ILDs. Therefore, this approach may be considered as a viable option for HI listeners with hearing aids to improve speech intelligibility in noise. Further, BEG has been well studied in hearing aid users but very little is known about it in cochlear implants (CIs) recipients. This group relies solely on ILDs to localize sounds as well as to understand speech in noise. As a consequence, CI recipients may benefit the most from any improvement in ILD processing. However, before any ILD enhancement method should be considered, it needed to be clarified if CI recipients are at all able to take advantage of BEG cues. The results of the fourth study revealed that at least some CI recipients can utilize BEG to some extent.