Beskrivelse
While a level increase typically gives better speech intelligibility (SI) at low presentation levels, it can lead to decreased SI at high levels. Termed rollover (RO), this suprathreshold phenomenon has the potential to affect aided outcome.Literature findings regarding RO occurrence are inconsistent. Some studies found RO at moderate levels already, whereas others observed RO at high levels only. Differences in test procedures, such as the use of speech materials with or without semantic context information, could be responsible for this (Fereczkowski & Neher, 2023). So far, guidelines for detecting and compensating for RO with hearing aids are sparse. One study compared linear amplification with two non-linear amplification conditions and observed comparable RO for the three of them (Shanks et al., 2002).
Here, we present results from ongoing research that investigates methods for detecting and compensating for RO. On the diagnostic side, we performed SI measurements using different speech materials and presentation levels with hearing-impaired listeners who were tested with linear amplification (N = 2 x 20). We found clear RO at moderate-to-high levels, regardless of whether speech materials with or without semantic context were used. Further, RO magnitude was comparable for the two types of speech materials.
On the hearing aid side, we present results for a fitting strategy for preventing RO in individual listeners. Termed sweet-spot compression, this strategy aims to place speech in that area of a given individual’s performance-intensity function where both intelligibility and listening comfort are high. In a proof-of-concept study, we showed that, in contrast to an NAL-NL1 reference condition, sweet-spot compression prevented RO and ensured better outcome at above-conversational levels (Fereczkowski et al., 2024). At low levels, however, listeners preferred the reference condition. Here, we present follow-up work that developed and evaluated sweet-spot compression further (N = 25). More specifically, the focus is on determining the individual sweet-spot area and on improving outcome at low levels.
Fereczkowski M & Neher T (2023). Semantic context can mask intelligibility declines at above-conversational speech levels in normal-hearing listeners. J Speech Lang Hear Res, 66(6), 2177-2183
Fereczkowski M, Sanchez-Lopez R H, Christiansen S & Neher T (2024). Amplitude compression for preventing rollover at above-conversational speech levels. Trends Hear, 28, DOI: 10.1177/23312165231224597
Shanks JE, Wilson RH, Larson V & Williams D (2002). Speech recognition performance of patients with sensorineural hearing loss under unaided and aided conditions using linear and compression hearing aids. Ear Hear, 23(4), 280-290
| Periode | 19. aug. 2024 → 25. aug. 2024 |
|---|---|
| Begivenhedstitel | International Hearing-Aid Seminar & International Hearing-Aid Research Conference |
| Begivenhedstype | Konference |
| Placering | Lake Tahoe, USA, CaliforniaVis på kort |
| Grad af anerkendelse | International |