J Am Acad Audiol 24:649–659 (2013)

The Effects of Noise Reduction Technologies on the Acceptance of Background Noise DOI: 10.3766/jaaa.24.8.2 Kristy Jones Lowery* Patrick N. Plyler*

Abstract Background: Directional microphones (D-Mics) and digital noise reduction (DNR) algorithms are used in hearing aids to reduce the negative effects of background noise on performance. Directional microphones attenuate sounds arriving from anywhere other than the front of the listener while DNR attenuates sounds with physical characteristics of noise. Although both noise reduction technologies are currently available in hearing aids, it is unclear if the use of these technologies in isolation or together affects acceptance of noise and/or preference for the end user when used in various types of background noise. Purpose: The purpose of the research was to determine the effects of D-Mic, DNR, or the combination of D-Mic and DNR on acceptance of noise and preference when listening in various types of background noise. Research Design: An experimental study in which subjects were exposed to a repeated measures design was utilized. Study Sample: Thirty adult listeners with mild sloping to moderately severe sensorineural hearing loss participated (mean age 67 yr). Data Collection and Analysis: Acceptable noise levels (ANLs) were obtained using no noise reduction technologies, D-Mic only, DNR only, and the combination of the two technologies (Combo) for three different background noises (single-talker speech, speech-shaped noise, and multitalker babble) for each listener. In addition, preference rankings of the noise reduction technologies were obtained within each background noise (1 5 best, 3 5 worst). Results: ANL values were significantly better for each noise reduction technology than baseline; and benefit increased significantly from DNR to D-Mic to Combo. Listeners with higher (worse) baseline ANLs received more benefit from noise reduction technologies than listeners with lower (better) baseline ANLs. Neither ANL values nor ANL benefit values were significantly affected by background noise type; however, ANL benefit with D-Mic and Combo was similar when speech-like noise was present while ANL benefit was greatest for Combo when speech spectrum noise was present. Listeners preferred the hearing aid settings that resulted in the best ANL value. Conclusion: Noise reduction technologies improved ANL for each noise type, and the amount of improvement was related to the baseline ANL value. Improving an ANL with noise reduction technologies is noticeable to listeners, at least when examined in this laboratory setting, and listeners prefer noise reduction technologies that improved their ability to accept noise. Key Words: Acceptable noise level, hearing aids Abbreviations: ANL 5 acceptable noise level; BNL 5 background noise level; Combo 5 combination; D-Mic 5 directional microphone; DNR 5 digital noise reduction; HINT 5 Hearing in Noise Test; MCL 5 most comfortable level; SRT 5 speech recognition threshold

H

earing loss affects an estimated 36 million people in the United States (Donahue et al, 2010). With no corrective treatment for a vast major-

ity of those afflicted, the most viable treatment option is hearing aids. In addition to amplifying desired sounds such as speech, hearing aids also amplify undesired

*Department of Audiology and Speech Pathology, University of Tennessee, Knoxville, TN Patrick N. Plyler, Ph.D., 578 South Stadium Hall, Knoxville, TN 37996-0740; E-mail: [email protected] Poster presentation at AudiologyNOW! 2009, Dallas, TX.

649

Journal of the American Academy of Audiology/Volume 24, Number 8, 2013

sounds such as background noise. Not surprising, one of the most common complaints of persons fitted with hearing aids involves listening to speech in the presence of background noise (Plomp, 1978; Dubno et al, 1984; Festen and Plomp, 1990; Souza and Turner, 1994; Needleman and Crandell, 1995; Killion, 1997). In order to meet the goal of communication restoration, advancements in hearing aid technology have focused on aiding speech while concomitantly addressing various background noises found bothersome to hearing aid users. Two such technologies designed to reduce the effects of background noise are directional microphones and digital noise reduction algorithms. The goal of directional microphone (D-Mic) technology is to attenuate sounds arriving at the hearing aid microphone from anywhere other than the front of the listener. In order to gain maximum benefit from directional technology, the hearing aid wearer should face the sound source of interest while undesired sounds are located behind them (Ricketts et al, 2003). Directional benefit is dependent upon several factors such as the number and location of speakers; the type, level, and distance of the noise source; the reverberation characteristics of the environment; and the amount of low-frequency compensation provided (Amlani, 2001). Digital noise reduction (DNR) technology works on the principle that physical characteristics of speech differ from physical characteristics of most noise-like stimuli, allowing for gain reductions of the noise-like input. While the implementation of DNR varies across manufacturers, most base a determination of gain reduction on analysis of the following aspects of the signal or environment: signal-to-noise ratio, input level, amplitude modulation frequency, and amplitude modulation depth of the signal. These components can be assessed independently or in various combinations to establish rules regarding how much and in what channels gain reduction should occur (Bentler and Chiou, 2006). Although the goal of both D-Mic and DNR technology is to improve performance in noise, recent research suggests that speech intelligibility may not influence a listeners’ preference for a hearing aid and may not be the best predictor of a successful hearing aid fitting (Nabelek et al, 2004, 2006). Acceptable noise level (ANL) was first introduced by Nabelek et al (1991) in an attempt to quantify the amount of background noise a listener is willing to accept while listening to speech. The premise of the ANL measure is that a person’s willingness to listen in noise may be more important than their ability to understand in noise for successful use of hearing aids (Nabelek et al, 1991). Acceptance of noise is measured as the difference between speech presented at the individual’s most comfortable listening level (MCL) and the highest background noise level (BNL) that is acceptable while listening to and following a speech sample (ANL 5 MCL 2 BNL). Consequently,

650

listeners with low ANL values accept more background noise when listening to speech while listeners with high ANL values accept less background noise when listening to speech. Research has shown that unaided ANL values serve as accurate predictors of hearing aid use. For example, Nabelek et al (2004, 2006) assessed hearing aid use patterns with a questionnaire. Instead of defining hearing aid use by the number of hours used per day, successful hearing aid users (full-time) were defined as wearing their hearing aids whenever they needed them, and unsuccessful hearing aid users were defined as wearing their hearing aids only occasionally (part-time) or not at all (nonuser). Successful hearing aid users accepted higher levels of background noise (i.e., have smaller ANLs) than unsuccessful hearing aid users, regardless of their ability to understand speech in noise (Nabelek et al, 2004, 2006). If technologies designed to reduce the effects of background noise can improve an ANL, the probability that individuals with high unaided ANL values could become successful hearing aid users could improve as well, thereby providing additional justification for utilizing certain noise reduction technologies. To date, limited research has been conducted investigating the effects of these technologies on the ANL. Freyaldenhoven et al (2005) investigated the effects of directionality on the ANL. Forty listeners with impaired hearing that used hearing instruments equipped with both omnidirectional and directional microphone capabilities were examined. Because the hearing aids were fit independent of the study, hearing aid style, type of directional microphone, vent size, compression, or lowfrequency gain compensation were not controlled. Each listener’s ANL, front-to-back ratio (FBR), and masked speech recognition threshold (SRT) were obtained for each microphone mode. For each test, speech was presented to the listener from a loudspeaker at 0° and noise from 180°. Results indicated that the average benefit of directionality, expressed as a decline in ANL, was 3.5 dB. This benefit was comparable to the 3.6 dB benefit measured as an improvement in masked SRT. These results suggested that acceptance of noise could be improved when using directional microphones; however, it should be noted that the variability in directional benefit obtained was quite large (16 dB). The effects of digital noise reduction on the ANL were investigated by Mueller et al (2006). Twenty-two adults were each fitted binaurally with behind-the-ear widedynamic-range compression hearing aids with DNR processing (Siemens ACURIS Model S). Each listener’s speech intelligibility and ANL were assessed with DNR on and DNR off using the Hearing in Noise Test (HINT) stimuli. For both tests, speech and noise were presented from the same loudspeaker at 0°. Results revealed a significant improvement on the ANL (4.2 dB) for the DNR-on condition compared to the DNR-off condition;

ANL: Noise Reduction/Lowery and Plyler

however, HINT results were unaffected by the use of DNR. Furthermore, the HINT results were not correlated with the ANL results for either condition (DNR on or off). These findings suggested that DNR can significantly improve acceptance of background noise and therefore may result in improved ease of listening for hearing aid users (Mueller et al, 2006). Recent research investigated the effects of noise management algorithms on speech intelligibility in noise and acceptance of noise in 18 participants fitted binaurally with either behind-the-ear (Widex Inteo) or in-thecanal hearing aids (Widex Inteo IN-X) (Peeters et al, 2009). Each listener’s performance was assessed under four hearing aid conditions: omnidirectional microphone only, DNR only (Speech Enhancer), fixed directional microphone only, and fixed directional microphone with DNR. Speech understanding was assessed using the HINT; however, acceptance of noise was measured using the Connected Speech Test (CST) passages (Cox et al, 1987) and speech-shaped noise from the HINT. For both tests, speech was presented at 0° and noise was presented at 90, 180, and 270°. Results indicated that each noise management condition significantly improved performance on the HINT and the ANL when compared to the omnidirectional condition (Peeters et al, 2009). While the aforementioned research provided promising insight regarding the use of D-Mic and DNR technologies to improve acceptance of noise, a systematic, well-controlled evaluation of these features on the ANL remains needed, as the methodologies used by Freyaldenhoven et al (2005), Mueller et al (2006), and Peeters et al (2009) may have impacted the results obtained. For example, Freyaldenhoven et al (2005) tested participants using their personal hearing aids, which did not control for factors known to affect directional benefit such as hearing aid style, type of directional microphone, compression, or venting (Ricketts, 2000a, 2000b). As a result, the variability in ANL benefit with directionality was relatively large and ranged from 24 to 12 dB. Thus, it is unclear if factors known to affect directional benefit on speech intelligibility measures also affected the ANL. In addition, ANL measurements are typically conducted while listening to continuous speech in the presence of continuous background noise (Nabelek et al, 2006). In contrast, Mueller et al (2006) and Peeters et al (2009) each made ANL measures using stimuli and procedures that differed from Nabelek et al (2006). For example, Mueller et al (2006) and Peeters et al (2009) instructed listeners to make perceptual judgments while listening to interrupted speech samples (HINT sentences, CST passages, respectively) in the presence of continuous background noise (HINT noise). If listeners made perceptual judgments when listening between speech samples (in noise alone) instead of while the speech was presented

(as instructed), noise levels could have been reduced by the DNR system, thereby potentially improving the ANL. Moreover, Peeters et al (2009) used a modified methodology and provided different instructions than those established by Nabelek et al (2006) when measuring MCL and BNL. Interestingly, the absolute ANL ranged from 218 to 3 dB in the Peeters et al (2009) study as compared to 2 to 28 dB in a previous ANL study (Nabelek et al, 2006). Thus, it is unclear if DNR can improve an individual’s ANL when listening to continuous speech in the presence of continuous noise when using ANL procedures established by Nabelek et al (2006). Previous research has also indicated that ANL values are not affected by the type of background noise used (Nabelek et al, 1991); however, the effect of background noise type on ANL benefit with directionality and/or digital noise reduction has not been examined. The amount of directionality provided by any directional microphone varies as a function of frequency as directivity index values are greater for low frequencies than high frequencies (Ricketts et al, 2005) and may be further impacted by venting and/or low frequency gain compensation (Freyaldenhoven et al, 2006). Moreover, the amount of noise reduction provided by any DNR algorithm varies as a function of noise type as DNR is more effective for steady-state noises than noises containing speech (Mueller and Ricketts, 2005). Therefore, both spectral and temporal differences that exist between various types of background noises could affect the functioning of the directional microphones and/or DNR systems, thereby potentially affecting the ANL. Importantly, the effect of ANL benefit with directionality and/or digital noise reduction on listener preference has not been evaluated. Although research suggests that successful hearing aid users accept higher levels of background noise (i.e., have smaller ANLs) than unsuccessful hearing aid users, what remains unclear is if improving an individual’s ANL with directionality and/or digital noise reduction is actually preferred by a hearing aid user or not. If listener preference corresponds to the condition that provides the lowest (best) ANL, attempting to improve an ANL with technology would clearly be a worthy goal. Therefore, the purpose of this experiment was to determine the effects of various noise reduction technologies on the ANL. The following research questions were addressed: 1. Is acceptance of noise affected when using noise reduction technologies (D-Mic, DNR, D-Mic 1 DNR)? 2. Is acceptance of noise affected by noise type when using noise reduction technologies (D-Mic, DNR, D-Mic 1 DNR)? 3. Is acceptance of noise related to subjective preference?

651

Journal of the American Academy of Audiology/Volume 24, Number 8, 2013

METHODS Participants Thirty adult listeners with sensorineural hearing impairment were recruited to participate in this experiment. Twenty-four males and six females with an average age of 65.5 (24–84) yr participated. Twenty-three of the participants were current, binaural hearing aid users while seven of the participants reported never using hearing aids. A power analysis computed with SPSS software was used to determine the number of listeners. The group size of 30 was estimated by a two-way fixed effect analysis of variance power analysis using a significance level of 0.05, power at 0.90, and effect size at 0.4. Listeners for the experiment were selected from the University of Tennessee Hearing and Speech Center as well as the Knoxville community. The criteria for inclusion included: (1) sensorineural hearing impairment with no more than a 15 dB difference in pure tone thresholds at any octave frequency from 250 through 8000 Hz between ears; (2) normal appearance of ear canal and pinna; (3) no air-bone gaps greater than 10 dB. Mean air-conduction hearing thresholds and standard deviations for left and right ears for the participants are shown in Figure 1. All qualification and experimental testing was conducted in a sound-treated examination room (Industrial Acoustic) with ambient noise levels suitable for testing with ears uncovered (American National Standards Institute [ANSI], 1999). Hearing Instruments Each participant was fitted binaurally with Siemens, Artis 2 S/VC digital behind-the-ear hearing instruments (Siemens Hearing Instruments Inc., Piscataway, NJ) using foam Comply tips (Hearing Components Inc., Oakdale, MN). The same two hearing instruments were used for each participant. The hearing aids employed 12 channel wide dynamic range compression with multiple memory capabilities, multichannel directional microphones, feedback cancellation, digital noise reduction, expansion, wind suppression, and volume controls. The hearing aids also allowed for continuous electromagnetic transmission between the instruments, ensuring that both hearing aids were operating in the same program/setting during testing. Prior to testing, measures were conducted using the Knowles Electronic Mannequin for Auditory Research to ensure that the features under investigation performed as expected. The hearing aids were initially programmed using proprietary fitting algorithm software, Siemens CONNEXX 5.0 version 1 (Siemens Hearing Instruments Inc.). Wind suppression and volume controls were disabled for the duration of testing; however, expansion and feedback cancellation remained enabled as their functioning did

652

Figure 1. Mean hearing thresholds and standard deviations for left and right ears for participants.

not serve to interfere with the features under investigation. Probe microphone measures were then conducted to verify match to National Acoustic Laboratories— Nonlinear 1 (NAL-NL1) targets (Byrne et al, 2001) for each participant in this baseline setting (omnidirectional with DNR deactivated), and baseline ANL measures were obtained as well. Following baseline measurements, the three memories in each hearing instrument were programmed randomly for each participant: (1) directional microphone activated only (D-Mic); (2) digital noise reduction activated only (DNR); and (3) directional microphone and digital noise reduction activated simultaneously (Combo). Thus, each memory had identical fitting parameters except for the respective activation/deactivation of the D-Mic and/ or DNR features. The hearing instruments employed the same DNR algorithm investigated by Mueller et al (2006), which allowed for direct comparison of results. The particular system utilized two different types of DNR algorithms, one modulation based and one an adaptive fast-acting system, much like Wiener filter technology (Hamacher et al, 2005). The systems, described extensively by Hamacher and colleagues (2005), operated simultaneously and independently in all 12 channels of the aids. The modulation-based algorithm analyzes the spectrum of the envelope in order to attenuate frequency components with very low signal-to-noise ratios whereas the adaptive fast acting system employs a 10 msec filter to track the signal envelope of each channel to provide intersyllabic noise reduction. Test Materials ANL values were measured using the Nabelek et al (2006) procedure with running speech recorded by a male talker as the primary stimulus (Arizona Travelogue, Cosmos Inc.). Three separate types of background noises were used for each memory: (1) a single male talker using

ANL: Noise Reduction/Lowery and Plyler

a recording of the Ipsilateral Competing Message from the Synthetic Sentence Identification with Ipsilateral Competing Message test (Speaks and Jerger, 1965); (2) 12talker speech babble (Revised SPIN recorded by Cosmos Inc.; Bilger et al, 1984); and (3) speech-shaped noise from the Hearing in Noise Test (HINT). The features under investigation are affected by the spectral and temporal properties of the background noise; therefore, the noise types chosen for this study were spectrally and temporally different from one another yet commonly encountered by hearing instrument users. (Note: the same speech stimuli and multitalker babble used in this study may currently be obtained from Frye Electronics.) Procedures ANL Measurements A randomized testing schedule was generated for each participant to determine the order in which memories and noise conditions would be evaluated. All speech and background noise stimuli were produced by a compact-disc player and routed through a twochannel diagnostic audiometer (GSI-61) calibrated to ANSI standards (ANSI, 2010) to ear-level loudspeakers located in a sound-treated booth. The speech signal was presented at 0°, and the background noise was presented at 180° in order to maximize the D-Mic effects. Participants were seated 1 m from the loudspeakers. Prior to data collection, participants were given oral and written instructions for the ANL procedure (Appendix 1). The intensity of the stimuli was manipulated in 5 dB steps initially and in 2 dB steps when selecting the final loudness level that was “most comfortable.” Once the participant’s MCL was established, background noise was introduced in the predetermined order set forth by the randomized experimental schedule. The background noise was introduced at 30 dB HL, as suggested by Nabelek et al (2004), and adjusted to the participant’s background noise level (BNL). BNL was defined as the level of background noise that could be accepted, without becoming tense or tired, while listening to speech. Each participant’s ANL was measured twice for each experimental condition (4 hearing aid settings 3 3 noises 3 2 5 24 total measurements), and the average of the two measures served as the ANL value for each listener for the given condition.

Preference Following the completion of ANL testing, each participant was asked to subjectively evaluate each memory. Participants were given oral and written instructions (Appendix 2). The speech signal was presented at 0°, and the background noise was presented at 180° at levels corresponding to each participant’s baseline (D-Mic and DNR deactivated) ANL for the respective noise condition. For example, if a participant’s baseline ANL resulted from an MCL of 58 dB HL and a BNL of 38 dB HL for the single-talker noise, the speech was presented at 58 dB HL, and the single-talker noise was presented at 38 dB HL. After being given adequate time to listen to the speech and noise in each program, participants were asked to rank the three memories, from 1 to 3, with 1 being the best and 3 being the worst. This ranking procedure was repeated for all three noise types and resulted in nine rankings for each participant. At the completion of the experiment, each participant was asked to indicate which memory they preferred overall to determine if the acceptance of noise was related to subjective preference. RESULTS ANL Measurements Acceptable noise level results were averaged across the 30 participants for each experimental condition (Table 1). The first purpose of the present study was to determine if acceptance of noise was affected when using noise reduction technologies. Consequently, ANL values were converted to benefit scores for the 30 participants to determine if acceptance of noise improved for the experimental conditions examined. Benefit was determined by subtracting the ANL value obtained from the baseline ANL value for the given condition. For example, a participant with a baseline ANL of 20 dB and a D-Mic ANL of 5 dB would have an ANL benefit of 15 dB for the given noise condition. Thus, positive values represent improvements in the ANL when using the noise reduction technology. Nine benefit scores were calculated for each participant. Benefit scores were then averaged within each experimental condition across the 30 participants (Fig. 2). The mean benefit score across noise type for each memory was as follows: DNR, 3.28 (213 to 17); D-Mic, 5.3 (28 to 19); and Combo, 7.01 (28 to 21).

Table 1. Mean, Range, and Standard Deviation of ANLs Averaged across Noise Type for the 30 Participants Noise Type Single Talker Spectrum Babble

Baseline

DNR

D-Mic

Combo

14.23, (22–35), 8.38 15.33, (22–31), 8.33 13.30, (26–32), 8.01

11.67, (2–26), 6.51 10.1, (0–25), 6.78 11.27, (21–29), 7.42

8.50, (25–26), 7.19 9.87, (21–25), 6.11 8.60, (27–24), 8.0

7.47, (24–24), 7.87 7.03, (22–24), 6.32 7.17, (22–19), 5.83

653

Journal of the American Academy of Audiology/Volume 24, Number 8, 2013

Figure 2. Mean ANL benefit results for each hearing aid setting and background noise condition. ANL benefit was determined by subtracting the ANL value obtained in a given condition from the baseline ANL value. Standard deviations are shown.

A two-way analysis of variance with repeated measures was performed to evaluate the effect of noise reduction technologies and background noise type on ANL benefit. The dependent variable was ANL benefit value. The within-subject factors were noise reduction technology (D-Mic, DNR, Combo) and background noise type (single-talker, twelve-talker babble, speech-shaped noise). The analysis revealed a significant noise reduction technology effect [F(2,58) 5 16.599, p , 0.01, partial h2 5 0.364, Ω 5 1.000] as well as a significant noise reduction technology 3 background noise type interaction, [F(4,116) 5 2.911, p , 0.05, partial h2 5 0.091, Ω 5 0.770]; however, the background noise type main effect was not significant [F(2,58) 5 2.682, p . 0.05, partial h2 5 0.085, Ω 5 0.512]. Paired samples t-tests were conducted to further investigate the noise reduction technology main effect. All comparisons were significant controlling for familywise error rate across tests at the .05 level, using the Holm’s sequential Bonferroni procedure (Table 2). These results indicate that listeners received significantly more benefit in the Combo program than in the D-Mic or DNR programs, and significantly more benefit in the D-Mic program than in the DNR program. Paired samples t-tests were also conducted to further investigate the noise reduction technology 3 background noise type interaction controlling for familywise error rate across tests at the .05 level, using the Holm’s sequential Bonferroni procedure within noise reduction technology condition (Table 3). Results indicated that

ANL benefit for D-Mic and Combo was significantly greater than ANL benefit with DNR for both the singletalker and multitalker babble noise conditions. For the speech spectrum noise, results indicated that ANL benefit for Combo was significantly greater than ANL benefit for both DNR and D-Mic. These results suggested that D-Mic and Combo provided more ANL benefit for background noise containing speech, but Combo provided more ANL benefit for non-speech-like background noise. Benefit data were averaged across the noise conditions for baseline, D-Mic, DNR, and Combo. Correlations were conducted to determine if a relationship existed between baseline ANL and the ANL benefit received from each noise reduction technology (Fig. 3). Correlations were significant for D-Mic (r 5 .508, p , 0.01), DNR (r 5 .556, p , 0.05), and Combo (r 5 .592, p , 0.01). These results suggested the relationship between baseline ANL and ANL benefit was strong (Cohen, 1988) for each noise reduction technology as listeners with higher (worse) baseline ANLs received more benefit from noise reduction technologies than listeners with lower (better) baseline ANLs. Preference Rankings Each participant was asked to rank his or her preferred hearing aid memory (1 5 best, 3 5 worst) while listening to speech and noise at levels corresponding to baseline values for each of the three background noises. The number of times each technology was preferred (ranking of 1) was calculated for each noise condition (Fig. 4). As there were 30 participants, the maximum number of “best” rankings a technology could receive would be 30 while the minimum number of “best” rankings a technology could receive would be 0. Three onesample x2 tests were conducted to assess the effect of noise reduction technology on rankings for the three background noise types. The results were significant for single talker [x2 (2, N 5 30) 5 7.80, p , 0.05], speech-shaped noise [x2 (2, N 5 30) 5 18.60, p , 0.05], and twelve-talker babble [x2 (2, N 5 30) 5 12.80, p , 0.05]. Follow-up testing for the single-talker noise condition indicated that the proportion of top rankings for both D-Mic and Combo were significantly greater than the proportion for DNR [x2 (1, N 5 12) 5 8.00, p , 0.05] [x2 (1,

Table 2. Results of the Paired Samples t-Tests for the Noise Reduction Technology Main Effect Controlling for Familywise Error Rate across the Tests at the .05 Level, Using the Holm’s Sequential Bonferroni Procedure Pair D-Mic, DNR D-Mic, Combo DNR, Combo

654

Mean Difference

Standard Deviation

t

df

Adjusted p

2.02 21.77 23.79

4.25 2.84 3.59

2.61 23.41 25.78

29 29 29

,0.01 ,0.01 ,0.01

ANL: Noise Reduction/Lowery and Plyler

Table 3. Results of the Paired Samples t-Tests for the Noise Reduction Technology by Background Noise Type Interaction Controlling for Familywise Error Rate across the Tests at the .05 Level, Using the Holm’s Sequential Bonferroni Procedure Noise Type Single Talker

Spectrum

Babble

Pair

Mean Difference

Standard Deviation

t

df

Adjusted p

D-Mic, DNR D-Mic, Combo DNR, Combo D-Mic, DNR D-Mic, Combo DNR, Combo D-Mic, DNR D-Mic, Combo DNR, Combo

3.16 21.03 24.2 0.23 22.83 23.06 2.66 21.43 24.10

5.39 3.59 4.89 4.71 3.83 4.60 5.06 4.73 4.70

3.21 21.54 24.70 0.27 24.04 23.65 2.88 21.65 24.77

29 29 29 29 29 29 29 29 29

,0.01 0.12 ,0.01 0.78 ,0.01 ,0.01 ,0.01 0.10 ,0.01

N 5 15) 5 5.40, p , 0.05]; however, the proportion of top rankings for D-Mic did not differ significantly from the proportion for Combo [x2 (1, N 5 28) 5 .333, p . 0.05]. For the speech-shaped noise condition, the proportion of top rankings for Combo was significantly greater than the proportion for both D-Mic [x2 (1, N 5 24) 5 13.5, p , 0.05] and DNR [x2 (1, N 5 27) 5 8.33, p , 0.05]; however, the proportion of top rankings for D-Mic and DNR were not significantly different [x2 (1, N 5 9) 5 1.00, p . 0.05]. For the twelve-talker babble noise condition, the proportion of top rankings for both D-Mic and Combo were significantly greater than the proportion for DNR [x2 (1, N 5 12) 5 5.33, p , 0.05] [x2 (1, N 5 20) 5 12.80, p , 0.05]; however, the proportion of top rankings for D-Mic did not differ significantly from the proportion for Combo [x2 (1, N 5 28) 5 2.29, p . 0.05]. These results suggested that D-Mic and Combo were ranked as best for background noise containing speech, but Combo was ranked as best for non-speech-like background noise. Overall Preference At the completion of all experimental testing, each participant was asked to indicate the memory they preferred overall (Fig. 5). A one-sample x2 test was conducted to assess overall preference. The results of the test were significant, x2 (2, N 5 30) 5 8.60, p , 0.05. The proportion of listeners that preferred Combo (p 5 .53) was greater than the hypothesized proportion of .33, whereas the proportion of listeners that preferred DNR (p 5 .10) was less than the hypothesized portion of .33. The proportion of listeners that preferred D-Mic (p 5 .36) did not differ from the hypothesized proportion of .33. Follow-up testing indicated that the proportion of listeners preferring Combo did not significantly differ from the proportion of listeners preferring D-Mic [x2 (1, N 5 28) 5 .926, p . 0.05]; however, the proportion of listeners preferring Combo and the proportion of listeners preferring D-Mic were significantly greater than those preferring DNR [x2 (1, N 5 19) 5 8.895, p , 0.05] [x2 (1, N 5 14) 5 4.571, p , 0.05].

DISCUSSION ANL Procedure The goal of the present study was to determine if acceptance of noise was affected when using noise reduction technology in hearing aids in the presence of different types of background noise, as measured by ANL. Listeners yielded lower (improved) ANL scores relative to baseline when using noise reduction technologies. This ANL benefit was evident with digital noise reduction technology, directional microphone technology, and the combination of the two technologies. Furthermore, the amount of ANL benefit differed with the respective technologies, with benefit increasing significantly from DNR to D-Mic to Combo. Directionality and ANL Results from this study revealed some difference in ANL benefit relative to previous studies investigating ANL and directionality. In comparing the present data to that of Freyaldenhoven et al (2005), the mean ANL

Figure 3. Correlation results between the baseline ANL and ANL benefit with each hearing aid setting averaged across each background noise condition. The dashed line represents the trendline for the DNR setting.

655

Journal of the American Academy of Audiology/Volume 24, Number 8, 2013

Figure 4. Number of times each hearing aid setting was ranked as the best setting for each background noise condition. Possible results ranged from 30 (best) to 0 (worst).

benefit was examined for the babble noise only. Those data revealed an ANL benefit of 4.7 dB with D-Mic, which, compared to the 3.5 dB benefit reported by Freyaldenhoven et al (2005), revealed an added benefit of 1.2 dB. Because the present data revealed no significant difference in ANL benefit scores between noise types, comparing the present data collapsed across noise types for the directional microphone program also provides an accurate comparison. Collapsed across noise type, the mean ANL benefit is 5.3 dB, which, compared to the Freyaldenhoven et al (2005) data, yields an added benefit of 1.8 dB. As Freyaldenhoven and colleagues (2005) did not control for factors known to affect directional benefit such as hearing aid style, type of directional microphone, vent size, or compression, the difference seen in the present study may be attributed to those factors. However, the variability noted with directional microphones in the presence of babble in the present study (28 to 16) was similar to that reported by Freyaldenhoven (24 to 12), suggesting that even if factors known to affect directionality are controlled for, individuals yield large variability in the amount of ANL benefit received with D-Mic technology. It should be noted that while a large range in ANL benefit was observed with D-Mic in the presence of speech babble, only 13% of participants received less benefit with D-Mic compared to baseline ANL scores. Further, 77% of participants received some benefit (.0 dB), while 10% of participants’ ANL scores remained at baseline. Previous research has indicated that ANL values are not affected by the type of background noise used, with the exception of music (Nabelek et al, 1991); however, the effect of background noise type on ANL benefit with directionality had not been examined to date. Prior to testing, the frequency response of each background noise used in the present study was visually examined and deemed to be spectrally different from one another. Despite the difference in frequency response, no signifi-

656

cant difference between the noises was observed in D-Mic ANL scores relative to baseline ANL scores. This lack of difference could be attributed to the noises not being different enough from one another to yield a difference in ANL score. Another explanation may lie in the measurement tool used to assess “directional benefit.” Previous research investigating directionality, specifically directionality as a function of frequency, did so using objective measures such as speech intelligibility (Ricketts et al, 2005). Because previous research has suggested ANL is not correlated to objective measures (Nabelek et al, 1991; Freyaldenhoven et al, 2006; Mueller et al, 2006), perhaps no difference was seen in the amount of ANL benefit because listeners were asked to perform a task seemingly unrelated to speech intelligibility. Therefore, a direct comparison should not be made between ANL benefit scores and scores of “directional benefit” obtained with speech intelligibility tasks. Further, while these data are in good agreement with previous data and suggest that D-Mic provides an average of 5.3 dB of ANL improvement; clinicians should be warned that considerable variability exists from patient to patient. Digital Noise Reduction and ANL The present data were also compared to data observed by Mueller et al (2006). Collapsing the present data across noise type for the DNR program yielded a mean ANL benefit of 3.3 dB, which is 0.9 dB lower than the 4.2 dB benefit reported by Mueller et al (2006). The DNR program for speech-shaped noise alone yielded an ANL benefit of 5.2 dB. Despite the methodological differences used in the two studies, similar benefit scores were obtained for DNR in the presence of speech-shaped noise. These results suggest that DNR, as it was similarly implemented in the two studies, can improve an individual’s ANL when listening to both continuous

Figure 5. Number of participants who preferred each hearing aid setting overall at the conclusion of the study. Possible results ranged from 30 (best) to 0 (worst).

ANL: Noise Reduction/Lowery and Plyler

and discontinuous speech in the presence of continuous noise. While these results reinforce the suggestion that DNR can result in improved ease of listening for speech in noise, it should be noted that the amount of ANL benefit will largely depend on the type of background noise present. The mean ANL benefit across noise types for DNR was nearly 2 dB lower (worse) than that for speech-shaped noise alone. This difference in ANL benefit is due to the significant difference noted in ANL benefit scores for speech-shaped noise relative to either single talker or speech babble. As DNR technology differentially amplifies speech and noise based on their physical characteristics and temporal differences, it is not surprising that significant differences were observed in ANL benefit among “speech-like” and “non-speech-like” background noises. These data also reinforce findings of Mueller and Ricketts (2005), who reported DNR to be more effective for steady-state noises than noises containing speech. Directionality 1 Digital Noise Reduction and ANL Previous research on noise reduction technology and ANL measures has focused on investigating D-Mic and DNR independently. What remained unclear was how the combination of technologies affected ANL. The current ANL benefit data revealed listeners’ ANL scores to be significantly better (lower) using the combination of technologies than either D-Mic or DNR alone, suggesting an additive benefit. Previous studies have revealed D-Mic to positively affect intelligibility, DNR to make no impact either positive or negative, and the combination of the technologies to yield results similar to those seen by D-Mic (Walden et al, 2000; Ricketts and Hornsby, 2005), thereby suggesting no additive benefit. While no difference in intelligibility was reported for DNR, both studies revealed a listener preference for DNR over D-Mic. While previous benefit studies have used objective intelligibility tasks as their measurement tool for benefit, it appears that whatever factor listeners used in assessing their preference for DNR in previous “objective benefit” studies was also a factor in the assessment of their own acceptance of noise. Interestingly, the correlational data correspond well with the ANL benefit data in that the DNR yields both the smallest benefit score and the weakest correlation, while Combo yields the largest benefit score and the strongest correlation. These data therefore suggest that individuals with larger baseline ANLs will receive greater ANL benefit from the noise reduction technologies under study than those individuals with smaller baseline ANLs. Further, the amount of benefit is correlated to the specific noise reduction technology used, with Combo clearly yielding the best opportunity for acceptance of noise. As such, according to the regression analysis performed by Nabelek et al (2006), an individual

Table 4. Mean ANL Benefit, Top Rankings, and Overall Preference Collapsed across Noise Type for the 30 Participants for Each Technology ANL Benefit (dB) Top Rankings Overall Preference

DNR

D-Mic

Combo

3.2 3.6 3

5.3 9.4 11

7.1 17 16

with a high unaided ANL could benefit greatly from being fit with a Combo of technology, which may increase their probability of success with hearing aids. Subjective Procedure Although the present study revealed results comparable to previous research in that ANL scores were significantly lower (better) with both D-Mic and DNR relative to baseline, what remained unclear was how this improvement would affect listener preference. Subjective preference data revealed that noise reduction technology did affect listener preference. In terms of preference and rankings data, listeners revealed the ability to detect some difference between the technologies under investigation. While a clear winner did not emerge between D-Mic and Combo in either background noise containing speech-like noise, it was evident that listeners preferred both D-Mic and Combo relative to DNR in these situations, suggesting that listeners were able to detect a difference between the memories with directional microphones (D-Mic and Combo) and the memory without directional microphones (DNR) in the presence of speech-like noise. Further, it is also evident that listeners preferred Combo for non-speech-like noise relative to either D-Mic or DNR, suggesting some additive benefit of D-Mic and DNR in the presence of non-speech-like noise. These data reflect the ANL benefit trends in that listeners preferred memories that provided the most acceptance of background noise within a given noise condition. The overall preference data follow the same trend as the seen in the aforementioned subjective rankings data as well as seen in the ANL benefit scores in that listeners preferred D-Mic and Combo significantly more than DNR. Overall, in reviewing the preference data along with the ANL benefit data, it is evident that improving an ANL with hearing aid technology is noticeable to listeners, at least when examined in a laboratory setting. For example, results from ANL benefit, rankings, and overall preference were collapsed across noise type for each hearing aid setting (Table 4). The ANL benefit increased by approximately 2 dB as the hearing aid settings changed from DNR to D-Mic to Combo. Similarly, changing the hearing aid settings from DNR, to D-Mic, to Combo resulted in an increase in the number of top rankings and an increase in overall preference as well.

657

Journal of the American Academy of Audiology/Volume 24, Number 8, 2013

These results suggest that listeners prefer conditions in which they are able to accept more noise relative to listening conditions in which they accept less noise. As such, the combination of technologies appears to be an effective method of managing background noise that is both quantifiably and qualitatively noticed by listeners. Conclusion The primary purpose of the present study was to determine if noise reduction technology affected acceptance of noise in the presence of different types of background noise. Results suggest that noise reduction technologies improve ANL in the presence of single-talker, speechshaped, and babble noise. Results further suggest that the amount of improvement depends upon an individual’s baseline ANL score, with the larger (worse) scores receiving more benefit from technology than smaller (better) baseline scores. In addition, the type of noise reduction technology employed as well as the type of background noise present affect the amount of benefit received, with Combo and D-Mic providing more benefit than DNR in the presence of speech-like background noise, and Combo providing more benefit than DNR and D-Mic in the presence of non-speech-like background noise. Also of interest was to determine if noise reduction technology affected subjective preference scores. Results suggest that ANL benefit impacts subjective preference insomuch as listeners prefer the noise reduction technologies that yielded the most improvement in terms of noise acceptance within a given listening condition. It should be noted that while these data are promising in terms of improving a listener’s acceptance of noise and perhaps their success with amplification, further investigation is warranted regarding how this ANL benefit translates into “real world success.” As these data were obtained in a laboratory, under ideal conditions, it is unclear if the findings will generalize to realworld success. In addition, these results can only be attributed to the testing set-up and hearing aid parameters implemented in the present study. Real-world settings rarely present speech at 0° azimuth only and noise at 180° azimuth only, as was investigated in the present study. As such, future research should investigate the effects of the technologies in more diffuse listening situations and perhaps with adaptive directional microphones and additional acclimatization time. Again, while these data are promising in terms of improving one’s acceptance of noise, the ultimate goal of such research is to help individuals with high ANLs become more successful hearing aid users. As such, a field-based investigation should explore whether an improvement in ANL, as provided by noise reduction technologies, produces a change in hearing aid use patterns. Also, in addition to a field-based investigation employing the technology explored in the present

658

research, additional ways to improve ANL scores should be explored. For example, FM systems, which improve signal-to-noise ratio, decrease distance between the sound source and listener, and overcome reverberation, should be investigated in terms of their ability to affect ANL independently of noise reduction technologies, as well as in combination with noise reduction technologies, as there may be an additive effect. In addition, auditory training as well as the role of the visual system may warrant investigation regarding their ability to affect an individual’s ability to accept noise. REFERENCES American National Standards Institute (ANSI). (1999) Maximum Ambient Noise Levels for Audiometric Test Rooms (ANSI S3. 1-1999). New York: American National Standards Institute. American National Standards Institute (ANSI). (2010) American National Standards Specification for Audiometers (ANSI S3. 62010). New York: American National Standards Institute. Amlani AM. (2001) Efficacy of directional microphone hearing aids: a meta-analytic perspective. J Am Acad Audiol 12(4):202–214. Bentler RA, Chiou LK. (2006) Digital noise reduction: an overview. Trends Amplif 10(2):67–82. Bilger RC, Nuetzel JM, Rabinowitz WM, Rzeczkowski C. (1984) Standardization of a test of speech perception in noise. J Speech Hear Res 27(1):32–48. Byrne D, Dillon H, Ching T, Katsch R, Keidser G. (2001) NAL-NL1 procedure for fitting nonlinear hearing aids: characteristics and comparisons with other procedures. J Am Acad Audiol 12(1):37–51. Cohen J. (1988) Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: L. Erlbaum. Cox RM, Alexander GC, Gilmore C. (1987) Development of the Connected Speech Test (CST). Ear Hear 8(5, Suppl.):119S–126S. Donahue AD, Dubno JR, Beck L. (2010) Guest editorial: accessible and affordable hearing health care for adults with mild to moderate hearing loss. Ear Hear 31(1):2–6. Dubno JR, Dirks DD, Morgan DE. (1984) Effects of age and mild hearing loss on speech recognition in noise. J Acoust Soc Am 76(1): 87–96. Festen JM, Plomp R. (1990) Effects of fluctuating noise and interfering speech on the speech-reception threshold for impaired and normal hearing. J Acoust Soc Am 88(4):1725–1736. Freyaldenhoven MC, Nabelek AK, Burchfield SB, Thelin JW. (2005) Acceptable noise level as a measure of directional hearing aid benefit. J Am Acad Audiol 16(4):228–236. Freyaldenhoven MC, Plyler PN, Thelin JW, Nabelek AK, Burchfield SB. (2006) The effects of venting and low-frequency gain compensation on performance in noise with directional hearing instruments. J Am Acad Audiol 17(3):168–178. Hamacher V, Chalupper J, Eggers J, et al. (2005) Signal processing in high-end hearing aids: state of the art, challenges, and future trends. EURASIP J Adv Signal Process 18:2915–2929. Killion M. (1997) The SIN report: circuits haven’t solved the hearing-in-noise problem. Hear J 50:28–32.

ANL: Noise Reduction/Lowery and Plyler

Mueller HG, Ricketts TA. (2005) Digital noise reduction: much ado about something? Hear J 58:10–17.

Ricketts TA. (2000a) Directivity quantification in hearing aids: fitting and measurement effects. Ear Hear 21(1):45–58.

Mueller HG, Weber J, Hornsby BW. (2006) The effects of digital noise reduction on the acceptance of background noise. Trends Amplif 10(2):83–93.

Ricketts TA. (2000b) Impact of noise source configuration on directional hearing aid benefit and performance. Ear Hear 21(3): 194–205.

Nabelek AK, Freyaldenhoven MC, Tampas JW, Burchfield SB, Muenchen RA. (2006) Acceptable noise level as a predictor of hearing aid use. J Am Acad Audiol 17(9):626–639.

Ricketts TA, Henry P, Gnewikow D. (2003) Full time directional versus user selectable microphone modes in hearing aids. Ear Hear 24(5):424–439.

Nabelek AK, Tampas JW, Burchfield SB. (2004) Comparison of speech perception in background noise with acceptance of background noise in aided and unaided conditions. J Speech Lang Hear Res 47(5):1001–1011.

Ricketts TA, Henry PP, Hornsby BW. (2005) Application of frequency importance functions to directivity for prediction of benefit in uniform fields. Ear Hear 26(5):473–486.

Nabelek AK, Tucker FM, Letowski TR. (1991) Toleration of background noises: relationship with patterns of hearing aid use by elderly persons. J Speech Hear Res 34(3):679–685. Needleman AR, Crandell CC. (1995) Speech recognition in noise by hearing-impaired and noise-masked normal-hearing listeners. J Am Acad Audiol 6(6):414–424. Peeters H, Kuk F, Lau CC, Keenan D. (2009) Subjective and objective evaluation of noise management algorithms. J Am Acad Audiol 20(2):89–98. Plomp R. (1978) Auditory handicap of hearing impairment and the limited benefit of hearing aids. J Acoust Soc Am 63(2):533–549.

Ricketts TA, Hornsby BW. (2005) Sound quality measures for speech in noise through a commercial hearing aid implementing digital noise reduction. J Am Acad Audiol 16(5): 270–277. Speaks C, Jerger J. (1965) Method for measurement of speech identification. J Speech Hear Res 8:185–194. Souza PE, Turner CW. (1994) Masking of speech in young and elderly listeners with hearing loss. J Speech Hear Res 37(3): 655–661. Walden BE, Surr RK, Cord MT, Edwards B, Olson L. (2000) Comparison of benefits provided by different hearing aid technologies. J Am Acad Audiol 11(10):540–560.

Appendix 1: ANL Instructions Provided to Each Participant Prior to Testing Instructions for establishing MCL: You will listen to a story through a loudspeaker. After a few moments, select the loudness of the story that is most comfortable for you, as if listening to a radio. Handheld buttons will allow you to make adjustments. First, turn the loudness up until it is too loud and then down until it is too soft. Finally, select the loudness level that is most comfortable for you. Instructions for establishing BNL: You will listen to the same story with background noise of several people talking at the same time. After you have listened to this for a few moments, select the level of background noise that is the most you would be willing to accept or “put up with” without becoming tense and tired while following the story. First, turn the noise up until it is too loud and then down until the story becomes very clear. Finally, adjust the noise (up and down) to the maximum noise level that you would be willing to “put up with” for a long time while following the story. Appendix 2: Subjective Rating Instructions Provided to Each Participant Prior to Testing You will be asked to listen to each memory under each of three noise conditions. For each noise condition, please rank each memory 1–3 (1 being the best, 3 being the worst) according to your preference for that particular memory/noise condition. Noise Condition 1 Memory 1 ______ Memory 2 ______ Memory 3 ______ Noise Condition 2 Memory 1 ______ Memory 2 ______ Memory 3 ______ Noise Condition 3 Memory 1 ______ Memory 2 ______ Memory 3 ______

659

Copyright of Journal of the American Academy of Audiology is the property of American Academy of Audiology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.