Effects of Digital Noise Reduction on Speech Perception for Children with Hearing Loss
Stelmachowicz, P., Lewis, D., Hoover, B., Nishi, K., McCreery, R., and Woods, W. (2010)
Because a great deal of everyday communication takes place in the presence of some level of background noise, hearing aid performance in noise is of interest to researchers, clinicians and hearing aid users. It is well established that directional microphones can improve signal-to-noise ratio (SNR) for adult hearing aid users as well as children (Valente et al. 1995; Gravel et al. 1999). It is generally accepted that Digital Noise Reduction (DNR) will not improve speech recognition in noise (Levitt et al. 1993; Bentler et al. 2008). Digital Noise Reduction has, however, resulted in improved overall sound quality judgments and decreased listening effort (Boymans & Dreschler, 2000; Walden et al., 2000; Sarampalis et al., 2009).
In noisy situations, adult listeners use a variety of cues to understand conversational speech, including visual cues, situational cues, semantic and grammatical context. Young children with limited language skills may not be able to take advantage of this information and may rely more on acoustic cues. Indeed, most studies show that children require better SNRs than adults (Blandy & Lutman, 2005; Jamieson et al. 2004).
For hearing-impaired children, hearing aids are more than a tool for the recognition of speech, they facilitate speech and language acquisition and development. As the authors of the current study pointed out, “amplification must facilitate the development of early auditory skills, laying the foundation for the extraction of regularities in the speech signal and the development of language.” Therefore, improving access to speech is of particular importance for young hearing aid users. Conversely, it is also must be determined that DNR or directional processing is not degrading the speech signal.
The purpose of the present study was to determine the effect of DNR on children’s perception of nonsense syllables, words and sentences in the presence of noise. Sixteen children with mild to moderately severe hearing loss participated in the study. Subjects were divided into two groups: 5-7 year olds and 8-10 year olds. The authors chose these age groups to evaluate the effect of development on the perception of speech stimuli with varying levels of context. Subjects were fitted with binaural behind-the-ear hearing aids with DNR and amplitude compression. Directional microphones were not activated. Hearing aids were programmed to DSL 5.0 targets and settings were verified with real-ear measurements.
The children were presented with speech stimuli mixed with speech-shaped noise at SNRs of 0dB, +5dB and +10dB. Three levels of context were represented:
- VCV (vowel-consonant-vowel) nonsense syllables, 15 consonants combined with /a/
- Monosyllabic words from the Phonetically Balanced Kindergarten List (PBK – Haskins 1949)
- Meaningful sentences with three key words each (Bench et al. 1979)
Data analysis revealed that noise reduction did not have a significant positive or negative effect on performance. There was no significant main effect for context, but not surprisingly, post hoc testing revealed that scores for both age groups were higher for sentences than they were for both nonsense syllables and monosyllables. Also not surprisingly, performance improved with increases in SNR for all types of speech stimuli. There was a significant effect of age, with older subjects demonstrating better overall performance than younger subjects. There was no interaction between age and noise reduction, indicating that the use of noise reduction did not affect performance of younger and older subjects differently. There was no interaction between age group and context, indicating that both age groups benefitted from context equivalently.
The authors observed a great deal of variability among subjects, especially the younger group. Though noise reduction did not significantly affect performance overall, the authors found that more than half of the younger subjects demonstrated poorer recognition of words in the DNR-on condition. The most common consonant confusions were: /f/ for /t/, /g/ for /d/, and /b/ for /v/, suggesting that voicing information was perceived correctly but place and manner of articulation were not easily distinguished. This finding is in agreement with previous results reported by Jamieson et al (1995) who found that DNR processing resulted in either no improvement or a slight decrement in performance and that consonant place of articulation was particularly affected. Granted, there are several cues that affect consonant perception and slight decrements in the acoustic representation of a consonant may be offset by the availability of other cues. For example, though /f/ and /t/ may be difficult to discriminate, a participant in face to face conversation benefits from visual cues to help identify these consonants. Still, the opportunity exists to further study the effect of noise reduction on consonant perception, with adult and pediatric subjects.
Despite the minimal effect of noise reduction on speech recognition, all listeners in Jamieson’s 1995 study reported a strong preference for DNR processing when hearing continuous speech in a variety of listening environments. This leads to an important consideration regarding the use of noise reduction processing in hearing aids for children. Although the current investigation did not address listening preference, previous studies with adults have often shown positive effects of noise reduction processing on listening effort and sound quality. The current authors suggested that if this were also the case for children, it could improve attentiveness and increase “time on task” in difficult listening situations. This is an interesting hypothesis, since attention and focus is essential for understanding speech in noise and many hearing-impaired children may demonstrate attention deficits.
Audiologists working with pediatric patients should consider noise reduction settings carefully. Although there were no statistically significant effects of noise reduction on speech perception in this study, decreases in word recognition scores for younger children in the DNR-on condition is a concern and warrants further study. The authors point out that a child’s ability to recognize and understand speech requires ongoing, consistent auditory experiences. Previous use of amplification, age of identification and consistency of hearing aid use may have influenced the results of this study and may affect success with DNR processing in general. The effect of degree of hearing loss should also be considered, as it is possible that individuals with severe hearing losses could be adversely affected by even small decrements in speech information resulting from DNR processing.
Clinically, an important highlight of this study is the fact that individual performance among children is highly variable. Digital Noise Reduction has the potential to ease listening, but may compromise clarity of speech. And directional microphones may improve access to speech, but also risk compromising speech audibility for off-axis talkers. These considerations suggest that some advanced features should be reserved for older children and specific environments. Among that older population, there may be some inclination to allow manual adjustment of hearing aid settings. However, Ricketts and Galster (2008) correctly point out that children cannot be expected to adjust manual directionality controls reliably. This ultimately results in a fitting rationale that avoids the fitting of some advanced features or allows them to function automatically, with the assumption that they will only be active in the appropriate situations and “do no harm” in regard to speech recognition.
Further study of the perceptual effects of noise reduction and subjective preferences in children is needed. The possibility remains that DNR may offer hearing-impaired children other benefits such as improved attention and comfort in noise, possibly leading to increased satisfaction and compliance from pediatric patients.
Bench, J., Kowal, A., & Bamford, J. (1979). The BKB sentence lists for partially-hearing children. British Journal of Audiology 13, 108-112.
Bentler, R., Wu, Y.H., Kettel, J. (2008). Digital noise reduction: Outcomes from laboratory and field studies. International Journal of Audiology 47, 447-460.
Blandy, S. & Lutman, M. (2005). Hearing threshold levels and speech recognition in noise in 7-year-olds. International Journal of Audiology 44, 435-443.
Boymans, M., & Dreschler, W.A. (2000). Field trials using a digital hearing aid with active noise reduction and dual-microphone directionality. Audiology 39, 260-268.
Gravel, J.S., Fausel, N., Liskow, C. (1999). Children’s speech recognition in noise using omnidirectional and dual microphone hearing aid technology. Ear and Hearing 20, 1-11.
Haskins, H.A. (1949). A phonetically balanced test of speech discrimination for children. Master’s thesis, Northwestern University, Evanston, IL.
Jamieson, D.G., Kranjc, G., Yu, K. (2004). Speech intelligibility of young school-aged children in the presence of real-life classroom noise. Journal of the American Academy of Audiology 15, 508-517.
Levitt, H., Bakke, M., Kates, J. (1993). Signal processing for hearing impairment. Scandanavian Audiology Supplement 38, 7-19.
Ricketts, T.A. & Galster, J. (2008). Head angle and elevation in classroom environments: implications for amplification. Journal of Speech, Language and Hearing Research 15, 516-525.
Sarampalis, A., Kalluri, S., Edwards, B., & Hafter, E. (2009). Objective measures of listening effort: Effects of background noise and noise reduction. Journal of Speech Language and Hearing Research, 52, 1230-1240.
Valente, M., Fabry, D., Potts, L.G. (1995). Recognition of speech in noise with hearing aids using dual microphones. Journal of the American Academy of Audiology 6, 440-449.
Walden, B.E., Surr, R.K., Cord, M.T. (2000). Comparison of benefits provided by different hearing aid technologies. Journal of the American Academy of Audiology 11, 540-560.