International Journal of Audiology

ISSN: 1499-2027 (Print) 1708-8186 (Online) Journal homepage: http://www.tandfonline.com/loi/iija20

Enhanced speech perception in noise and cortical auditory evoked potentials in professional musicians Kiriana Meha-Bettison, Mridula Sharma, Ronny K. Ibrahim & Pragati Rao Mandikal Vasuki To cite this article: Kiriana Meha-Bettison, Mridula Sharma, Ronny K. Ibrahim & Pragati Rao Mandikal Vasuki (2017): Enhanced speech perception in noise and cortical auditory evoked potentials in professional musicians, International Journal of Audiology, DOI: 10.1080/14992027.2017.1380850 To link to this article: http://dx.doi.org/10.1080/14992027.2017.1380850

View supplementary material

Published online: 03 Oct 2017.

Submit your article to this journal

Article views: 12

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iija20 Download by: [Australian Catholic University]

Date: 07 October 2017, At: 15:29

International Journal of Audiology 2017; Early Online: 1–13

Original Article

Enhanced speech perception in noise and cortical auditory evoked potentials in professional musicians

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

Kiriana Meha-Bettison1, Mridula Sharma2,3, Ronny K. Ibrahim2,3, and Pragati Rao Mandikal Vasuki2,4 1

Australian Hearing, The Australian Hearing Hub, Macquarie University, Sydney, Australia, 2Department of Linguistics, The Australian Hearing Hub, Macquarie University, Sydney, Australia, 3The HEARing CRC, Audiology, Hearing and Speech Sciences, The University of Melbourne, Melbourne, Australia, and 4Audiology Research, Starkey Hearing Research Centre, Berkeley, USA

Abstract Objective: The current research investigated whether professional musicians outperformed non-musicians on auditory processing and speech-in-noise perception as assessed using behavioural and electrophysiological tasks. Design: Spectro-temporal processing skills were assessed using a psychoacoustic test battery. Speech-in-noise perception was measured using the Listening in Spatialised Noise – Sentences (LiSN-S) test and Cortical Auditory Evoked Potentials (CAEPs) recorded to the speech syllable/da/presented in quiet and in 8-talker babble noise at 0, 5, and 10 dB signal-to-noise ratios (SNRs). Study sample: Ten professional musicians and 10 non-musicians participated in this study. Results: Musicians significantly outperformed non-musicians in the frequency discrimination task and low-cue condition of the LiSN-S test. Musicians’ N1 amplitude showed no difference between 5 dB and 0 dB SNR conditions while non-musicians showed significantly lower N1 amplitude at 0 dB SNR compared to 5 dB SNR. Brain-behaviour correlation for musicians showed a significant association between CAEPs at 5 dB SNR and the low-cue condition of the LiSN-S test at 30–70 ms. Time–frequency analysis indicated musicians had significantly higher alpha power desynchronisation in the 0 dB SNR condition indicating involvement of attention. Conclusions: Through the use of behavioural and electrophysiological data, the results provide converging evidence for improved speech recognition in noise in musicians.

Key Words: Musicians, auditory processing, speech-in-noise perception, cortical auditory evoked potentials, alpha, theta, oscillations, event related spectral perturbations

Introduction The last decade has seen an increasing number of papers gathering evidence on the topic of enhanced auditory and/or cognitive skills in musicians (Pantev et al. 2001; Janata, Tillmann, and Bharucha 2002; Schneider et al. 2002; Shahin et al. 2003; Kraus and Chandrasekaran 2010; Pallesen et al. 2010). There is also evidence that plastic changes, such as differences in the auditory cortex of adult musicians, could possibly be attributed to the effects of longterm musical training and regular practice/auditory training (Pantev et al. 2001; Schneider et al. 2002; Pantev and Herholz 2011). In general, musicianship has been linked to superior performance on a number of auditory skills such as segregation of simultaneously occurring sounds (Zendel and Alain 2009), categorical perception (Bidelman et al. 2014) as well as listening in an acoustically

challenging environment such as reverberation (Bidelman and Krishnan 2010). Studies have reported that musicians have enhanced auditory skills, specifically pitch discrimination, as compared to nonmusicians (Kishon-Rabin et al. 2001; Micheyl et al. 2006; Bidelman, Krishnan, and Gandour 2011; Mandikal Vasuki et al. 2016). While the literature supports a difference on specific auditory skills between musicians and non-musicians, there is equally compelling evidence for lack of superior performance of musicians when compared to non-musicians on all auditory tasks. For instance, in a study that investigated seven different temporal tasks, musicians outperformed non-musicians on some of the temporal tasks such as rhythm discrimination, auditory fusion and temporal discrimination but not on duration percept

Correspondence: A/Prof Mridula Sharma, PhD, Audiology Program, Department of Linguistics, Macquarie University, Room 1.609, S2.6, NSW 2109, Australia. Tel: +61 2 9850 4863. E-mail: [email protected]

(Received 9 March 2017; revised 29 August 2017; accepted 11 September 2017) ISSN 1499-2027 print/ISSN 1708-8186 online ß 2017 British Society of Audiology, International Society of Audiology, and Nordic Audiological Society DOI: 10.1080/14992027.2017.1380850

2

K. Meha-Bettison et al. Abbreviations

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

ABR CAEP EEG ERP ERSP GFP HINT LiSN-S SAM SES SNR SRT

Auditory brainstem response cortical auditory evoked potential electroencephalography event related potential event related spectral perturbation global field power hearing in noise test listening in spatialised noise – sentences sinusoidal amplitude modulation socioeconomic status signal to noise ratio speech reception threshold

(Rammsayer and Altenmu¨ller 2006). Similarly, Mandikal Vasuki et al. (2016) reported that musicians, when tested on six different auditory processing skills, showed a significant advantage over nonmusicians only on the pitch discrimination task. A growing body of literature has investigated whether musicians may have an advantage over non-musicians in speech in noise perception (see review by Coffey, Mogilever, and Zatorre [2017]). An explanation of why musicians may have an advantage in speech in noise perception is cogently articulated in the OPERA hypothesis (Patel 2011). The hypothesis states that there is a significant overlap in the brain networks required for the neural encoding of music and language. Additionally, adaptive plasticity in speech processing networks may be driven through the precision of music processing, emotional engagement of music, repetitive nature of musical practice and engagement of focussed attention during musical tasks. However, the current evidence for musicians’ advantage in speech-in-noise processing is mixed (Ruggles, Freyman, and Oxenham 2014; Swaminathan et al. 2015; Clayton et al. 2016). Parbery-Clark et al. (2009b) found significant differences on speech recognition in noise across musicians and non-musicians when measured on the Quick Speech in Noise test and Hearing-in-Noise Test (HINT), while Ruggles, Freyman, and Oxenham (2014) using the same speech-in-noise tasks found no differences across the groups. Two other studies by Mandikal Vasuki et al. (2016) and Yeend et al. (2017) reported no significant differences between professional musicians and non-musicians on another speech-innoise task (Listening-in-Spatialised Noise – Sentence test; LiSN-S) that manipulates spatial and pitch cues. Previous research which specifically manipulated spatial locations and intelligibility of background noise found that musicians elicited better thresholds when speech and noise were not co-located (Swaminathan et al. 2015; Clayton et al. 2016). On the other hand, Parbery-Clark et al. (2009b) found that musicians showed an advantage on the HINT when there were no spatial cues available. In summary, there is some evidence that musicians have superior speech-in-noise processing. One reason for the differences across the reported research may be that different studies have applied varied test batteries. Another reason may be due to individual variability, as behavioural adaptive tasks require active participation and consistent active attention from all participants (Coffey, Mogilever, and Zatorre 2017). Interpreting behavioural data can, therefore be challenging, as it requires a response from the listener. Additionally, behaviour studies also do not help in identifying the underlying mechanisms that may be influencing superior behavioural performance.

Inclusion of electrophysiological measures in conjunction with behavioural measures helps overcome some of these challenges. Similar to reports on behavioural measures of auditory processing, there are reports that in some instances, musicians have larger amplitudes of evoked responses, indicating superior auditory processing when compared to non-musicians (Pantev et al. 2001; Shahin et al. 2003; Pantev and Herholz 2011). For instance, musicians have been reported to show enhanced neural encoding of vowel and resistance to reverberant noise (Bidelman and Krishnan 2010). Likewise, there is some evidence of enhanced cortical auditory evoked responses (responses such as N1-P2; MMN, ERAN) in adult musicians (Koelsch, Schmidt, and Kansok 2002; Shahin et al. 2003; Fujioka et al. 2004; Zendel and Alain 2009; Seppa¨nen et al. 2012). However, there is at least one paper that has found no differences to piano tones on auditory evoked fields between musicians and non-musicians (Lu¨tkenho¨ner, SeitherPreisler, and Seither 2006). In summary, there are dichotomous findings in the current literature on electrophysiological results similar to those for behavioural results. The current study investigated the performance of musicians and non-musicians on speech-in-noise and auditory processing tasks using a unique array of behavioural and electrophysiological measures. The electrophysiological measures included understanding the effect of varying level of noise on syllable percept at cortical level. Using time-domain analysis, we measured waveform latency and amplitude of the cortical auditory evoked potentials (CAEPs) which are considered to be indicators of decreased audibility, as morphology significantly degrades for all subjects when speech is presented in increasing levels of background noise (Billings et al. 2009, 2011; Sharma et al. 2014). Along with the traditional time-domain analysis, time-frequency analysis of electrophysiological data was undertaken to elucidate the neural oscillatory activity. Analysing this oscillatory activity in specific frequency bands is informative as different frequency components have been reported to have distinct functional roles in processing a signal (Klimesch 1999; Bhattacharya and Petsche 2005; Cohen 2017). For instance, alpha and theta oscillations may contribute to cognitive (attention) and memory aspects within a task (Klimesch 1999). In the current study, we were interested in the neural oscillatory activities across frequency bands, specifically the alpha band (frequencies between 8 and 12 Hz), at specific time-periods to determine if musician expertise is associated with enhanced cognitive abilities, specifically attention. In line with the OPERA hypothesis and based on the available studies (Shahin et al. 2003; Parbery-Clark et al. 2009b; Bidelman and Krishnan 2010; Bidelman et al. 2014), we hypothesised that the musicians’ neural encoding of speech would be similar to nonmusicians in quiet conditions but will show less degradation in noise especially at the most challenging noise levels, potentially due to their strengthened encoding of acoustic cues at the brainstem and cortical levels. Therefore, the overall aim of the current research was to investigate if musicians, when compared to non-musicians had: (a) superior performance on the behavioural speech in noise task, (b) superior performance on behavioural psychoacoustic auditory processing tasks; (c) larger neural responses on electrophysiological speech in noise task; and (d) increased involvement of alpha oscillations in the most adverse situation implying the role of non-auditory factors such as attention in listening in noise. The findings would be an initial step towards understanding the effect of musicianship and potential applications of music as a viable

Enhanced speech perception in noise and CAEPs in musicians intervention tool when working with populations with listening in noise difficulties.

Methods

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

Participants Twenty-three participants between the ages of 22 and 59 were recruited for participation in this study. All participants completed a questionnaire to assess previous musical and tonal language experience. The latter was a consideration as the effect of tonal language experience on speech perception in noise has been previously reported (Chandrasekaran, Krishnan, and Gandour 2009; Marie, Kujala, and Besson 2012). Twelve participants who selfidentified themselves as ‘‘professional musicians’’ were initially recruited. These participants reported drawing a significant portion of their income (430%) from playing music in a professional capacity over at least the past 3 years. Additional inclusionary criteria for musicians included commencing musical training at or prior to the age of 10 and a minimum of 15 h per week spent playing, practising or training in music. Two musician participants were excluded for having a significant hearing loss and one nonmusician was excluded for having a minor musical experience in childhood. The data reported here are from the final sample of 10 musicians and 10 non-musicians. Table 1 shows the musical training history of the musicians. There were no significant differences between the groups in age (U ¼ 42, p ¼ 0.545, medianM ¼ 51; medianNM ¼ 46) or gender (NM: 7 males, 3 females; NNM: 7 males, 3 females). Education and socio-economic status (SES) was assessed using an informal questionnaire in which the definition of SES and its various levels were included from the Australian taxation office website. While two of the musician participants declined to share their education and SES levels, when Mann–Whitney U was conducted on the available data there were no significant differences for the education level between the groups (U ¼ 25, p ¼ 0.155, NM ¼ 8; NNM ¼ 10). On the other hand, SES of the non-musicians was significantly higher than musicians [U ¼ 13, p ¼ 0.012, NM ¼ 8; NNM ¼ 10].

Peripheral hearing assessment All participants had their peripheral hearing assessed. Audiometric thresholds, at octave intervals between 250 and 8000 Hz including mid-frequencies 3000 and 6000 Hz, were measured. In addition, tympanograms, ipsilateral and contralateral acoustic reflexes and Distortion Product Otoacoustic Emissions (DPOAEs) were recorded bilaterally for each participant. All participants had pure tone

3

thresholds less than 20 dB HL at frequencies 250–2000 Hz and less than 45 dB HL at frequencies 3000–8000 Hz. For both groups, the average 3000–6000 Hz thresholds were less than 20 dB HL. However, three musicians and two non-musicians had thresholds between 25 and 45 dB HL for frequencies between 3000 and 6000 Hz. Mann–Whitney U was conducted for the averaged 3000– 6000 Hz thresholds for both groups and there was no significant difference across the groups (U ¼ 33.5, p ¼ 0.212; medianM ¼ 18.75; medianNM ¼ 6.67, NM ¼ NNM ¼ 10). In addition, no participant had an air-bone gap of greater than 15 dB HL at any frequency or any indication of middle ear dysfunction in either ear.

Psychoacoustic data Spectral and temporal processing were evaluated using a behavioural auditory processing test battery that assessed participants’ performance on pitch discrimination, gaps-in-noise (GIN) and sinusoidal amplitude modulation (SAM) at 4 and 64 Hz. All tasks were presented using the APEX 3 platform (Francart, Van Wieringen, and Wouters 2008). All stimuli were presented under headphones at a self-determined comfortable level. An adaptive three-alternative forced choice paradigm was employed for all tasks. Participants were presented with three stimuli, which consisted of two standard stimuli and one target variable signal presented in random order for each trial, each with an inter-stimulus interval of 500 ms (Peter et al. 2014). Participants were required to identify the target signal by clicking the box corresponding to the target. Thresholds were determined using a 2-down, 1-up tracking method to estimate the 70.7% correct point on the psychometric function (Levitt 1971). The behavioural auditory processing tasks are described in further detail in Table 2.

Speech-in-noise perception Speech-in-noise perception was behaviourally assessed with the LiSN-S, a sentence repetition task that measures speech reception thresholds for target sentences presented in competing background distractor speech. The test was administered on a PC under headphones (Sennheiser HD215) to create a spatial three-dimensional auditory environment and controlled using the Phonak LiSNS software (Cameron and Dillon 2007). The task comprised of four conditions in which spatial separation and vocal identity between the target and distracter voice was varied. These conditions included: (a) different voices at ±90 (high-cue SRT), (b) the same voice at ±90 , (c) different voices at 0 and (d) the same voice at 0 (low-cue SRT).

Table 1. Musician participants’ musical training history.

1 2 3 4 5 6 7 8 9 10

Age of onset of musical training

Years of training

Years playing professionally

Practice hrs/week

Primary instrument

Other instruments

7 3 9 9 7 5 8 7 8 9

43 43 46 40 20 54 44 46 16 45

22 30 33 26 3 40 30 30 5 32

15–20 40 25 25–30 25–30 15–20 30 20 24 15–20

Vocals Piano Bass guitar Guitar Drums Bass guitar Vocals Guitar Drums Vocals

Guitar N/A Guitar Congas Piano Guitar Keyboard Guitar Keyboard Drums Piano Violin Vocals Guitar Bass Keyboard Percussion

4

K. Meha-Bettison et al.

Table 2. Description of psychoacoustic auditory processing test procedures. Pitch discrimination

Gaps-in-noise detection

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

Temporal Modulation Transfer Function

Determines the smallest difference in frequency (in Hz) between two stimuli required for them to be distinguished as separate pitches. The standard stimulus was a 1000 Hz pure tone and the target stimulus was always higher in frequency. All tones had an equal duration of 200 ms with an onset and offset ramp of 10 ms. Trials continued until five reversals were obtained and the threshold corresponds to the mean of the final three reversals. Presentation and response collection was controlled by Apex 3 software (Francart, Van Wieringen, and Wouters, 2008). Determines the threshold (in ms) for detection of a short temporal gap contained within a broadband white noise. All stimuli was a 50–10 000 Hz broadband noise of 500 ms duration with an onset and offset ramp of 20 ms. The target signal included a silent gap with the onset corresponding to the midpoint of the noise. The duration of the gap for the initial trial was 10 ms and was adjusted for each successive trial until a total of eight reversals were obtained after a brief practice period. The threshold corresponds to the mean of the final eight reversals. Presentation of stimuli and response collection for this task was controlled by custom software developed by Dr Richard Baker (Baker et al., 2008). Determines thresholds for the detection of amplitude modulation as a function of modulation frequency. All stimuli were white noise, low pass filtered at 20 000 Hz with a duration of 500 ms and an onset and offset ramp of 20 ms. The standard stimulus was an un-modulated white noise and the target was the same noise with a sinusoidal amplitude modulation applied to the noise carrier at frequencies of both 4 and 64 Hz. The depth of amplitude modulation was adjusted for each successive trial until five reversals were obtained and thresholds correspond to the mean of the final three reversals. Presentation of stimuli and response collection for the TMTF was controlled by Apex 3 software (Francart, Van Wieringen, and Wouters, 2008). The modulation depth was expressed in dB (20 log m) where 0 dB indicates 100% modulation.

Target sentences were developed by a number of Australianregistered speech language pathologists who specialised in rehabilitation of children with hearing loss. The sentences were created in accordance with the criteria used for the development of original Bamford–Kowal–Bench (BKB) sentences and were deemed suitable for participants 4.6 years of age and older (Cameron and Dillon 2007)1. These sentences were spoken by an Australian female and were always presented at 0 azimuth. The noise or competing speech was also spoken in a female voice but the exact vocal identity and location along the azimuth was manipulated for each condition. In each condition, the participant would hear the distracter recite children’s stories continuously at a constant level of 55 dB SPL. On each trial, a 1000 kHz tone burst would alert the participant that the target sentence was about to be presented. Participants were required to repeat as many words as possible within each sentence and scoring was based on the number of words that were correctly repeated. The target speech was initially presented at a level of 62 dB SPL and the level then adjusted adaptively for each trial. When a participant accurately repeated more than 50% of the words, the level of the target sentence was decreased by the software. In the event that a participant scored less than 50%, the level was increased and the level would remain unchanged if the participant identified 50% of the words. For each condition, the task commenced with practice trials consisting of a minimum of five sentences and continuing until at least one reversal was obtained. Thresholds for each condition represent the SNR, which is the level of speech where the participant can repeat at least 50% of the words of a target sentence.

Electrophysiological data STIMULI The electrophysiological data was collected using the speech syllable /da/, spoken in isolation by an Australian female. This stimulus has a duration of 158ms and involves the ‘‘ah’’ vowel produced without an ‘‘r’’ sound, as in the pronunciation of the word ‘‘hard’’ in Australian English with a fundamental frequency of

238 Hz. The stimulus was perceptually identified as /da/by native Australian English listeners (Sharma et al. 2006). The /da/stimulus was presented in four conditions, in quiet and in presence of an 8-talker (Australian speakers) babble noise (Keidser et al. 2002) at three SNRs (10, 5 and 0 dB) with an interstimulus interval of 1100 ms. The speech stimulus was presented at a fixed level of 65 dB SPL and the level of the noise was adjusted corresponding to the SNR. Calibration of the speech stimuli (/da/and the background babble noise) was carried out using a sound level metre (SLM) Bru¨el and Kjaer (B & K) Type 2231, artificial ear Type 4152, 1-inch pressure microphone Type 4144, 2cc coupler BD 1038 and oscilloscope. The speech stimuli were calibrated in dB SPL (rms) using a microphone connected to a B & K measuring amplifier (linear weighting, impulse response) and an oscilloscope. Oscilloscope measurements of /da/in varying level of noise were undertaken to verify the SNRs. The background noise was 5 min long and set to loop. The four conditions (quiet, 10 dB, 5 dB and 0 dB SNR) were presented in a random order for each participant, in blocks where each condition was presented twice. Within each block, there were 300 stimulus presentations (2 blocks 150 presentations). Stimuli were presented binaurally through standard insert earphones designed specifically for use with NeuroscanÕ equipment. Both speech and noise stimuli was delivered by Neuroscan STIM2 (Compumedics).

CAEP

RECORDINGS

The CAEPs (gain 1k) were recorded in an acoustically and electrically shielded room. Participants were seated in a comfortable chair while they watched a muted movie with subtitles. Participants were requested to be as still as possible during the experiment. CAEPs were recorded using NeuroscanÕ Acquire 4.5 (Compumedics) from 27 electrode sites [F3, F1, FZ, F2, F4, FC3, FC1, FCZ, FC2, FC4, C3, C1, CZ, C2, C4, CP3, CP1, CPZ, CP2, CP4, P3, P1, PZ, P2, P4, M1 and M2]. These electrodes were positioned on an elastic cap (Easy Cap) using the International 10/20 EEG system.

Enhanced speech perception in noise and CAEPs in musicians Continuous electroencephalography (EEG) was recorded using the left mastoid as reference (M1 electrode) with online filtering of 0.01–100 Hz (Klem et al. 1999). Vertical eye movements were measured using electrodes placed above and below the right eye. Horizontal eye movements were measured using electrodes on the outer canthus of each eye. Electrode impedances were kept below 5 kO to ensure good electrode conductance.

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

DATA ANALYSIS Time domain EEG analysis. All CAEPs were analysed offline in NeuroscanÕ Edit 4.5 (Compumedics). The continuous EEG data were re-referenced to the average of the left (M1) and right (M2) mastoids. The re-referenced data was divided into 700 ms epochs, which ranged from 100 to 600 ms relative to the onset of the stimulus. Offline filtering was carried out with a pass band of 0.1–30 Hz to emphasise slow low frequency neural activity of the cortical response and all files were baseline corrected using the 100 ms mean value of the pre-stimulus interval. Artefact rejection criteria of ±50 mV were applied to remove extraneous physiological or other noise sources. For each SNR condition, the accepted CAEP epochs were averaged in the two blocks to obtain a single averaged waveform (4 waveforms per participant). The P1, N1 and P2 components were manually identified in each individual waveform at every electrode site. The first author identified the latencies for P1, N1 and P2 peaks such that in quiet P1 was the first positive peak between 30 and 100 ms, N1 was the first negativity between the 75–200 ms and P2 was the positivity following N1 and within the 150–250 ms (Martin, Tremblay, and Korczak 2008). In general, 10% of the peaks were randomly checked across the groups by the second and/or the fourth author. The P1–N1 and N1–P2 peak-to-peak amplitudes were calculated from absolute peak amplitudes. Global field power (GFP) was also calculated for the two groups for each SNR. The GFP was determined to quantify the total instantaneous global activity across the scalp as a function of time. The GFP is a reference-independent, maximal electric activity across the scalp at a certain time moment (Lehmann and Skrandies 1980). The GFP is the standard deviation of the waveforms for all scalp locations recorded for each SNR and each participant. The GFP provides the advantage that the results can be compared from the whole scalp (recorded location) rather than any one electrode location (Hamburger and Vd Burgt 1991).

5

corrected between 200 ms and 0 ms. Ocular artefact correction was performed using independent component analysis (ICA). Horizontal and vertical eye movements were identified visually for each subject and were excluded from the EEG reconstruction of the component space. In order to remove noisy trials, a variance rejection criterion was used. Trials with variances of more than 300 mV2 between 200 ms up to 800 ms were excluded from further analysis. For the purpose of frequency domain analysis, the accepted trials were transformed into time frequency domain using the wavelet transform, which utilised the Mortlet wavelet with a filter tap of five cycles to ensure a good time-frequency resolution. The time-frequency power spectrums were then converted into a relative power measure or better known as event related spectral perturbation (ERSP) with an average power baseline period of 200 ms to 0 ms relative to sound onset. Brain oscillatory activities represented by the ERSP were analysed to extract alpha (frequencies 8–12 Hz) oscillations.

Statistical analysis. Statistical analyses for the behavioural and time domain responses were performed using Statistica 12 (StatSoft Inc). Based on previous studies that reported the influence of age on musicianship (Zendel and Alain 2012; Parbery-Clark et al. 2012a, 2012b), we used age (meanM ¼ 46.90, SD ¼ 11.84; meanNM ¼ 44.80, SD ¼ 12.02) as a covariate. Similarly, to ensure that variability of hearing thresholds was also taken under consideration, average pure tone threshold between 3000 and 6000 Hz (meanM ¼ 17.58, SD ¼ 12.36; meanNM ¼ 10.50, SD ¼ 11.65) was also used as a covariate. The two groups were compared on various behavioural psychoacoustic and speech-innoise tasks using a multivariate analysis of covariance (MANCOVA). In addition, post-hoc comparisons were planned for the low-cue condition of LiSN-S. The latencies and amplitudes of the P1, N1 and P2 as well as the P1–N1 and N1–P2 complex components of the CAEPs recorded at each electrode site were also compared using MANCOVAs. As the main interest across the groups was at the worst SNRs of 5 and 0 dB, planned comparisons were employed for P1, N1 and P2 latency and amplitudes. Monte Carlo cluster-based permutations dependent regression slope t-test was performed across the groups to determine if there were any significant differences between musicians and nonmusicians in the lower alpha band (8–10 Hz).

Results Brain-behaviour correlation. Following procedures outlined in Mandikal Vasuki et al. (2017), a brain-behavioural correlation was performed to investigate the association between CAEPs in noise (10, 5 and 0 dB SNR) and low-cue SRT scores (LiSN-S). This was performed using BESA Statistics 2.0. The software corrects for multiple comparisons over a large group of electrodes through data clustering and permutation testing (Maris and Oostenveld 2007).

Time frequency EEG analysis. The offline time-frequency analysis of the EEG was performed using the open source Fieldtrip implemented in MATLAB (Oostenveld et al. 2011). The data were re-referenced to the average of the left (M1) and the right (M2) mastoids. The EEG trials were then divided into trials/epochs that started 600 ms prior to the auditory stimulus and ended 1100 ms after the sound onset (600 to 1100 ms). The epochs were baseline

Musicians outperformed non-musicians on pitch discrimination, amplitude modulation and in the most challenging condition of the behavioural speech-in-noise task. CAEP results showed N1 amplitude and P1 latency to be significantly different only at 0 dB SNR across the two groups.

Psychoacoustic data Musicians had significantly lower psychoacoustic thresholds than non-musicians on the pitch discrimination task (F[1, 16] ¼ 8.56, p ¼ 0.009) (Figure 1). Musicians significantly outperformed nonmusicians on the 4 Hz SAM task (Figure 2; F[1, 16] ¼ 5.78, p ¼ 0.029) but not the 64 Hz SAM task (F[1, 16] ¼ 3.79, p ¼ 0.069). There were no significant differences between the groups on the gap detection task (meanM 3.94 ± 0.64; meanNM 4.68 ± 0.98; F [1, 16] ¼ 3.97, p ¼ 0.064).

6

K. Meha-Bettison et al. NON-MUSICIANS

Speech-in-noise perception

Frequency Discrimination (log)

3.0 2.5 2.0

1.7

1.6

1.5 1.1

1.1 1.0

1.6

1.2 1.1 0.9

0.8

1.1

1.2

0.5

Correlation of behavioural results

0.0

Participants

Mean

Frequency Discrimination (log)

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

MUSICIANS 3.0 2.5 2.0 1.5 1.0 1.0 0.9 0.5 0.0

There were no main effects of LiSN-S and group (Figure 3; F[3, 48] ¼ 1.51, p ¼ 0.224). Planned comparison for the low-cue condition of LiSN-S test, however, showed a significant group effect (p ¼ 0.019). In other words, musicians outperformed nonmusicians on the most challenging condition of the LiSN-S test (same voice at 0 ).

0.9 0.6 0.6

1.0 0.7

0.6

0.7

0.4

0.4 Participants

Non-parametric analysis was undertaken due to small sample size in each group. Spearman rank correlation was undertaken for the musicians and non-musicians to investigate whether any of the LiSN-S subtasks were associated with performance on other auditory processing task such as pitch discrimination, GIN and SAM tasks. A significant correlation was obtained between the performance on pitch discrimination and the low-cue condition of LiSN-S task for musicians (rM ¼ 0.65, p50.05), while non-musicians showed no significant correlations between any of the auditory processing tasks and LiSN-S conditions.

Electrophysiological data Mean

Figure 1. The pitch discrimination threshold for each of the participants in two groups on log scale. The musician group showed significantly lower thresholds compared to the non-musicians when age and pure-tone average thresholds between 3000 and 6000 Hz were included as covariates.

Figure 4 shows/da/-evoked P1-N1-P2 responses recorded from Cz in quiet and at three different SNRs (10, 5 and 0) for both groups including standard deviation (light grey colour). In general, the participants in both groups showed robust responses and there was in general good consistency across subjects and electrode channels (Supplementary Figure S1), within each SNR and within each group. To overcome the problem of multiple comparisons over

Figure 2. Significant group differences for the thresholds of the 4 and 64 Hz modulation frequency across musician and non-musician groups when age and pure-tone average thresholds between 3000 and 6000 Hz were included as covariates.

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

Enhanced speech perception in noise and CAEPs in musicians

7

Figure 3. Group effects for LiSN-S conditions, where only low-cue task in musicians showed significantly lower SRT when age and puretone average thresholds between 3000 and 6000 Hz were included as covariates.

several electrode sites, an alpha level 0.002 was considered significant after Bonferroni correction [a’ ¼ 1  (1  a)1/k]. In general, both groups showed significant differences across the SNR such that in quiet, all peaks (P1–N1–P2) showed biggest amplitudes and shortest latencies compared to other SNRs across all electrodes. N1 latency across all electrodes was affected by noise in both groups and showed significant increase in latency as the noise level increased [F(3, 48) ¼ 11.06, p ¼ 0.0001]. There were no group effects. There was a significant main interaction between SNR and group for N1 amplitude across all electrodes [F(3, 48) ¼ 3.01, p ¼ 0.004]. Planned comparisons using t-tests across all electrodes revealed musicians showed no difference for N1 amplitude between the 5 dB SNR and 0 dB SNR conditions, while non-musicians showed a significant reduction in the N1 peak amplitude (p ¼ 0.0001) between 5 dB SNR and 0 dB SNR. There were no group or main effects for P1 latency and a planned comparison of the P1 latency component at 0 dB SNR revealed that musicians had significantly earlier latency than non-musicians across all electrodes (p ¼ 0.007). There were no other significant differences across the groups for P2 latency or P1–N1, N1–P2, P2 amplitude. The GFP did not show any significant difference across the groups for any of the SNRs (Supplementary Figure S2).

Brain-behavioral correlation Correlation analysis of the LiSN-S low-cue condition and CAEPs at different SNRs across all electrodes for musicians showed a significant cluster at frontal regions (p ¼ 0.04) for the time region 70–100 ms at 5 dB SNR (Figure 5). There were no other significant correlations for musicians or non-musicians.

Time–frequency analyses Figure 6 presents the relative change in percentage from the baseline for the non-musician and the musician groups. We performed a non-parametric independent sample t test at all SNRs to determine ERSP differences between musicians and nonmusicians. The test revealed that the alpha power desynchronisation (frequency 8–12 Hz) at 0 dB SNR was significantly higher for the musicians compared to the non-musician group at the fronto-central regions (p ¼ 0.012) (Figure 7).

Discussion Musicians have enhanced pitch discrimination, amplitude modulation and speech-in-noise perception The primary objective of this study was to investigate the musical advantage for speech-in-noise perception and auditory processing. The current findings are in line with our hypothesis and previous research that musicians show enhanced performance on frequency discrimination (Ruggles, Freyman, and Oxenham 2014; Mandikal Vasuki et al. 2016), SAM (Yeend et al. 2017) and speech perception in the presence of noise tasks (Parbery-Clark et al. 2009b). Musicians performed significantly better than non-musicians for only the low-cue SRT condition of the LiSN-S test. This is the most difficult condition, in which both the target and background speech signal share the same vocal identity and spatial location at 0 . Musicians and non-musicians performed similarly in all other conditions of the LiSN-S which involve spatial separation or differences in vocal identity between the target and background speech signals. Similar performance of both groups in spatially separated conditions was consistent with the previous study

K. Meha-Bettison et al.

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

8

Figure 4. Comparison of musician and non-musician average waveforms (bold line) at Cz with one standard deviation (grey) elicited in response to/da/in quiet, and in the presence of noise (SNR10, SNR5 and SNR0). All waveforms are on the same scale (top right waveform). Time (ms) is on the x-axis while amplitude (mV) is on the y-axis.

Enhanced speech perception in noise and CAEPs in musicians

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

by Parbery-Clark et al. (2009b) but inconsistent with other studies that indicate a musician advantage in spatially separated conditions such as Clayton et al. (2016) and Swaminathan et al. (2015). It is noteworthy that the LiSN-S test only evaluates speech in presence of noise at a wide spatial separation of 90 and 0 azimuths and only has one distractor. It would be useful to systematically evaluate if musicians have an advantage at finer spatial separations and the type of noise that is employed (Coffey, Mogilever, and Zatorre 2017). Interestingly, neither Mandikal Vasuki et al. (2016) nor Yeend et al. (2017) observed any difference in the low-cue LiSN-S between musicians and non-musicians. A potential reason for inconsistent results between these studies and the current study may

Figure 5. Topoplots showing cluster of electrodes where significant correlations were noted between LiSN-S low cure condition and CAEPs at 30–70 ms in the musician group.

9

be related to the age range and duration of musical training. In the current study, 8 out of 10 musicians have over 40 years of training experience while Mandikal Vasuki et al. had relatively younger musicians of which only two musicians had more than 40 years of musical experience. Yeend and colleagues recruited musicians with less than 8 years of music training to professional musicians. Future studies, which include groups of musicians with different ranges of musical training (e.g. 10 years, 15 years, 20 years etc.) may help to tease apart the influence of duration of musical training on speech in noise performance. Results from our psychoacoustic test battery revealed that musicians demonstrate both lower pitch discrimination thresholds and SAM thresholds at 4000 Hz than non-musicians, suggesting enhanced spectral and temporal envelope processing for musicians. Lower pitch discrimination thresholds is one of the most established differences between the groups (Kishon-Rabin et al. 2001; Tervaniemi et al. 2005). Superior temporal resolution in musicians has also been demonstrated despite the differences in the task used (Rammsayer and Altenmu¨ller 2006). Surprisingly, we did not observe a significant difference for gaps-in-noise thresholds across groups, however this is consistent with the findings of Mandikal Vasuki et al. (2016). Detection of sinusoidal amplitude modulation was significant in this group only for 4 Hz which is previously noted in other studies (Jain, Mohamed, and Kumar 2014; Yeend et al. 2017). There is also evidence of enhanced frequency following response (FFR) in musicians (Lee et al. 2009) which has been reported to correlate with SAM (Akhoun et al. 2008). It is surprising that dichotomous results were seen between the results of the current study and Mandikal Vasuki et al. (2016) even though similar paradigms were used. This discrepancy may again be related to the participant demographics in the two cohorts. For instance, in the current study, the participants were sampled across a wide age range and were trained in a variety of musical instruments (Table 1), and as previously identified most musicians had over 40 years or more of training. Nevertheless, the findings further

Figure 6. Event-Related Spectral Perturbation (ERSP) plots from electrodes C1, Cz, C2 for non-musician (row 1) and musician (row 2) group across the tested SNRs (columns: Quiet, 10, 5 and 0 dB). Red colour on the ERSP plots show a positive increase in brain oscillatory activity and blue colour shows a decrease in brain oscillatory activity. Only alpha desynchronisation is significantly higher in musicians at 0 dB SNR.

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

10

K. Meha-Bettison et al.

Figure 7. Alpha ERSP (shown by black dotted rectangles) for (a) non-musicians; and (b) musicians where significant difference between the two groups was seen (time 150–250 ms; p ¼ 0.015) while (c) shows statistical topographical map tests for the same time region at 0 dB SNR.

necessitate an in-depth evaluation of temporal perception in this population. Interestingly, significant correlations were obtained between performance on the pitch discrimination task and the low-cue condition of the LiSN-S in musicians. It is plausible that enhanced encoding of auditory stimuli (as measured through smaller pitch discrimination thresholds) may be contributing towards enhancements in speech in noise performance. As the current study only has 10 musicians, these findings should be investigated in a larger sample of musicians. The finding that musicians demonstrated significantly less waveform degradation than non-musicians (i.e. no significant difference between the amplitude of N1 between 5 dB and 0 dB SNR conditions) further supports the notion of enhanced encoding in musicians. Additionally, at 0 dB SNR, the musician group elicited significantly earlier P1 latency than the nonmusician cohort. Although noise has a negative effect on speech-in-noise perception for both groups alike, the impact appears to be smaller for musicians as they experience less degradation of the CAEP response with increasing levels of background noise. Our findings are in line with previous studies which show that musicians have a more robust encoding in noise at both cortical and brainstem level (Zendel et al. 2015; Parbery-Clark, Skoe, and Kraus 2009a; Bidelman and Krishnan 2010). Our finding is also consistent with studies that have found the differences may be more apparent between the musicians and non-musicians at low SNRs (Parbery-Clark, Skoe, and Kraus 2009a; Bidelman and Krishnan 2010). The brain-behaviour correlation showed a

significant cluster for the frontal electrodes for the musicians between 30–70ms at 5 dB SNR and not at 0 dB SNR. The involvement of the frontal electrodes at this time course may be indicative of differences in the P1–N1 response noted in the musicians. However, the lack of correlation for 0 dB SNR may be due to poor waveform morphology and consequently larger individual variability at this low SNR. In tandem with the time-domain analysis, the ERSP data also showed that noise appears to impact neural oscillations differently in the musicians and non-musicians. The musicians show alpha desynchronisation at 0 dB SNR that are significantly different between the two groups. Klimesch (1999) in his review paper noted alpha oscillations vary according to task complexity and task demands. In other words, in demanding conditions, alpha desynchronisation show an increase consistent with what we observed for musicians at 0 dB SNR. It could be that we are observing the ‘‘anticipatory attention’’ in musicians especially since the stimulus delivery was fixed at regular intervals (Klimesch 2012). At 0 dB SNR, the masking was effective, making the perception of/da/difficult especially for non-musicians as can be observed by the ERPs (Figure 4, Supplementary Figure S1). As suggested by Klimesch (2012), it is possible that the alpha desynchronisation in musicians at 0 dB SNR is a reflection of their ability to maintain the attention on the target information. While speculative, alpha power desynchronisation may be evidence of better selective attention skills in musicians. In other words, perhaps this cohort of musicians was more adept at listening to relevant and selectively ignoring the irrelevant signals even when not required to actively attend (Kaganovich et al. 2013).

Enhanced speech perception in noise and CAEPs in musicians

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

Explaining the musical advantage The current study provides neural and behavioural evidence of musicians’ advantage on some tasks but this was apparent only at the most difficult listening conditions. This may be because the long-term musical experience/training enhances the neural encoding of acoustic signals as proposed by the OPERA hypothesis (Patel 2011). The OPERA hypothesis states that ‘‘experience-dependent plasticity’’ occurs if five conditions are met, namely Overlap, Precision, Emotion, Repetition and Attention. The current data supports the OPERA hypothesis in that the musicians showed significantly better performance on pitch discrimination and amplitude modulation (4 Hz). The ‘‘overlap’’ aspect of OPERA implies that there should be a connection in the networks that process common features between music and speech. Both pitch discrimination and amplitude modulation are features that can change within music and are regarded imperative for music appreciation (Patel and Balaban 2000; McDermott and Oxenham 2008). Thus, presumably, ‘‘precise’’ encoding in musicians contributed to their superior performance on these tasks. Additionally, significant correlations were observed between the low-cue condition of LiSN-S and pitch discrimination in musicians. Taken together, the behaviour and electrophysiology results provide converging evidence of ‘‘precise’’ neural representation at various levels of auditory pathway which possibly assist during speech recognition in noise. An alternative or additional explanation for the differences between musicians and non-musicians may be related to enhanced selective and/or focussed attention in musicians. The OPERA hypothesis recognises the importance of focussed attention on ‘‘experience-dependent plasticity’’. Previous literature has linked increased amplitude of the N1 component of the CAEP to selective attention (Thornton, Harmer, and Lavoie 2007). Our finding of increased alpha desynchronisation in musicians at 0 dB SNR (Figures 6 and 7) supports this interpretation. Previous studies have reported that musicians have better attentional capacity and this could be the reason for superior ERP amplitudes even in the tasks where explicit attention is not required (Parbery-Clark, Skoe, and Kraus 2009a; Bidelman et al. 2014).

Conclusions, implications, limitations and future directions The differences between musicians and non-musicians on behavioural and electrophysiological measures add to the growing literature on experience-dependent plasticity in musicians. While the results are not conclusive of the underlying reasons for these differences, the current study provides tentative evidence of enhanced auditory processing as well as attentional processes in the current cohort of professional musicians. The finding that there are associations between musical expertise and speech perception in noise may have interesting implications. For example, clinical audiologists may need to consider the importance of musical experience in their interpretation of audiological assessments and management of cases. One limitation of the current study was the inclusion of only 10 musician participants with a wide age-range but with at least 40 years of musical experience. The participating musicians were somewhat heterogeneous in that they play variety of different instruments and not just one type of instrument. In future studies, it would be useful to investigate the auditory processing as measured by behavioural and electrophysiological tasks within a smaller age range, specific musical expertise and larger cohorts to explore the effect of the duration of musical training.

11

Additionally, whilst our findings support the model of experience-dependant neuroplasticity in musicians, the importance and exact role of musical training is unknown. In future studies, factors such as genetics, pre-disposition towards music, IQ, selective attention and working memory could also be controlled. Our data demonstrates that musicians exhibit an advantage in auditory processing, but fails to explain the mechanism underpinning such improvements. Future research must focus on determining where musicians derive their benefit, specifically addressing the potential role of selective attention and working memory, both of which could be integral to further understanding the musician advantage for speech-in-noise and auditory processing.

Acknowledgements The authors acknowledge the financial support of the HEARing CRC, established under the Cooperative Research Centres (CRC) Programme. The CRC Programme supports industry-led end-user driven research collaborations to address the major challenges facing Australia.

Declaration of interest: None of the authors have potential conflicts of interest to be disclosed.

Note 1. As described in Cameron and Dillon (2007) the LiSN sentences, similar to the BKB sentences, contain some Stage 2 (e.g. kick ball) and mostly Stage 3 or three-element clauses of subject, verb and object (e.g. daddy kick ball).

References Akhoun, I., S. Galle´go, A. Moulin, M. Me´nard, E. Veuillet, and C. BergerVachon. 2008. ‘‘The Temporal Relationship between Speech Auditory Brainstem Responses and the Acoustic Pattern of the Phoneme/BA/in Normal-Hearing Adults.’’ Clinical Neurophysiology 119: 922–933. Baker, R. J., D. Jayewardene, C. Sayle, and S. Saeed. 2008. ‘‘Failure to Find Asymmetry in Auditory Gap Detection.’’ Laterality 13: 1e21. Bhattacharya, J., and H. Petsche. 2005. ‘‘Phase Synchrony Analysis of EEG During Music Perception Reveals Changes in Functional Connectivity due to Musical Expertise.’’ Signal Processing 85: 2161–2177. Bidelman, G. M., and A. Krishnan. 2010. ‘‘Effects of Reverberation on Brainstem Representation of Speech in Musicians and Non-Musicians.’’ Brain Research 1355: 112–125. Bidelman, G. M., A. Krishnan, and J. T. Gandour. 2011. ‘‘Enhanced Brainstem Encoding Predicts Musicians’ Perceptual Advantages with Pitch.’’ European Journal of Neuroscience 33: 530–538. Bidelman, G. M., M. W. Weiss, S. Moreno, and C. Alain. 2014. ‘‘Coordinated Plasticity in Brainstem and Auditory Cortex Contributes to Enhanced Categorical Speech Perception in Musicians.’’ The European Journal of Neuroscience 40: 2662–2673. Billings, C. J., K. L. Tremblay, G. C. Stecker, and W. M. Tolin. 2009. ‘‘Human Evoked Cortical Activity to Signal-to-Noise Ratio and Absolute Signal Level.’’ Hearing Research 254: 15–24. Billings, C. J., K. O. Bennett, M. R. Molis, and M. R. Leek. 2011. ‘‘Cortical Encoding of Signals in Noise: Effects of Stimulus Type and Recording Paradigm.’’ Ear and Hearing 32: 53.

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

12

K. Meha-Bettison et al.

Cameron, S., and H. Dillon. 2007. ‘‘Development of the Listening in Spatialized Noise-Sentences Test (LISN-S).’’ Ear and Hearing 28: 196–211. Chandrasekaran, B., A. Krishnan, and J. T. Gandour. 2009. ‘‘Relative Influence of Musical and Linguistic Experience on Early Cortical Processing of Pitch Contours.’’ Brain and Language 108: 1–9. Clayton, K. K., J. Swaminathan, A. Yazdanbakhsh, J. Zuk, A. D. Patel, and G. Kidd, Jr. 2016. ‘‘Executive Function, Visual Attention and the Cocktail Party Problem in Musicians and Non-Musicians.’’ PLoS One 11: e0157638. Cohen, M. X. 2017. ‘‘Where does EEG Come From and What Does it mean?’’ Trends in Neurosciences 40: 208–218. Coffey, E. B., N. Mogilever, and R. J. Zatorre. 2017. ‘‘Speech-In-Noise Perception in Musicians: A Review.’’ Hearing Research 352: 49–69. Francart, T., A. Van Wieringen, and J. Wouters. 2008. ‘‘APEX 3: A MultiPurpose Test Platform for Auditory Psychophysical Experiments.’’ Journal of Neuroscience Methods 172: 283–293. Fujioka, T., L. J. Trainor, B. Ross, R. Kakigi, and C. Pantev. 2004. ‘‘Musical Training Enhances Automatic Encoding of Melodic Contour and Interval Structure.’’ Journal of Cognitive Neuroscience 16: 1010–1021. Hamburger, H. L., and M. A. Vd Burgt. 1991. ‘‘Global Field Power Measurement Versus Classical Method in the Determination of the Latency of Evoked Potential Components.’’ Brain Topography 3: 391–396. Jain, C., H. Mohamed, and A. U. Kumar. 2014. ‘‘Short-Term Musical Training and Pyschoacoustical Abilities.’’ Audiology Research 4: 102. Janata, P., B. Tillmann, and J. J. Bharucha. 2002. ‘‘Listening to Polyphonic Music Recruits Domain-General Attention and Working Memory Circuits.’’ Cognitive, Affective, & Behavioral Neuroscience 2: 121–140. Kaganovich, N., Kim, J. Herring, C. Schumaker, J. MacPherson, M. Weber ,and C. Fox. 2013. ‘‘Musicians Show General Enhancement of Complex Sound Encoding and Better Inhibition of Irrelevant Auditory Change in Music: An ERP Study.’’ European Journal of Neuroscience 37: 1295–1307. Keidser, G., T. Ching, H. Dillon, K. Agung, C. Brew, and S. Brewer. 2002. ‘‘The National Acoustic Laboratories (NAL) CDs of Speech and Noise for Hearing Aid Evaluation: Normative Data and Potential Applications.’’ Australian and New Zealand Journal of Audiology 24: 16. Kishon-Rabin, L., O. Amir, Y. Vexler, and Y. Zaltz. 2001. ‘‘Pitch Discrimination: Are Professional Musicians Better Than NonMusicians?’’ Journal of Basic and Clinical Physiology and Pharmacology 12: 125–144. Klem, G. H., H. O. Lu¨ders, H. Jasper, and C. Elger. 1999. ‘‘The Ten-Twenty Electrode System of the International Federation.’’ Electroencephalography Clinical Neurophysiology 52: 37–46. Klimesch, W. 1999. ‘‘EEG Alpha and Theta Oscillations Reflect Cognitive and Memory Performance: A Review and Analysis.’’ Brain Research. Brain Research Reviews 29: 169–195. Klimesch, W. 2012. ‘‘Alpha-Band Oscillations, Attention, and Controlled Access to Stored Information.’’ Trends in Cognitive Sciences 16: 606–617. Koelsch, S., B. H. Schmidt, and J. Kansok. 2002. ‘‘Effects of Musical Expertise on the Early Right Anterior Negativity: An EventRelated Brain Potential Study.’’ Psychophysiology 39: 657–663. Kraus, N., and B. Chandrasekaran. 2010. ‘‘Music Training for the Development of Auditory Skills.’’ Nature Reviews. Neuroscience 11: 599–605. Lee, K. M., E. Skoe, N. Kraus, and R. Ashley. 2009. ‘‘Selective Subcortical Enhancement of Musical Intervals in Musicians.’’ The Journal of Neuroscience 29: 5832–5840. Lehmann, D., and W. Skrandies. 1980. ‘‘Reference-Free Identification of Components of Checkerboard-Evoked Multichannel Potential Fields.’’ Electroencephalography and Clinical Neurophysiology 48: 609–621. Levitt, H. 1971. ‘‘Transformed UpDown Methods in Psychoacoustics.’’ The Journal of the Acoustical Society of America 49: 467–477.

Lu¨tkenho¨ner, B., A. Seither-Preisler, and S. Seither. 2006. ‘‘Piano Tones Evoke Stronger Magnetic Fields than Pure Tones or Noise, both in Musicians and Non-Musicians.’’ Neuroimage 30: 927–937. Mandikal Vasuki, P. R., M. Sharma, K. Demuth, and J. Arciuli. 2016. ‘‘Musicians’ Edge: A Comparison of Auditory Processing, Cognitive Abilities and Statistical Learning.’’ Hearing Research 342: 112–123. Mandikal Vasuki, P. R., M. Sharma, R. K. Ibrahim, and J. Arciuli. 2017. ‘‘Musicians’ Online Performance During Auditory and Visual Statistical Learning Tasks.’’ Frontiers in Human Neuroscience 11: 114. Marie, C., T. Kujala, and M. Besson. 2012. ‘‘Musical and Linguistic Expertise Influence Pre-Attentive and Attentive Processing of NonSpeech Sounds.’’ Cortex 48: 447–457. Maris, E., and R. Oostenveld. 2007. ‘‘Nonparametric Statistical Testing of EEG- and MEG-data.’’ Journal of Neuroscience Methods 164: 177–190. Martin, B. A., K. L. Tremblay, and P. Korczak. 2008. ‘‘Speech Evoked Potentials: From the Laboratory to the Clinic.’’ Ear and Hearing 29: 285–313. McDermott, J. H., and A. J. Oxenham. 2008. ‘‘Music Perception, Pitch, and the Auditory System.’’ Current Opinion in Neurobiology 18: 452–463. Micheyl, C., K. Delhommeau, X. Perrot, and A. J. Oxenham. 2006. ‘‘Influence of Musical and Psychoacoustical Training on Pitch Discrimination.’’ Hearing Research 219: 36–47. Oostenveld, R., P. Fries, E. Maris, and J.-M. Schoffelen. 2011. ‘‘FieldTrip: Open Source Software for Advanced Analysis of MEG, EEG, and Invasive Electrophysiological Data.’’ Computational Intelligence and Neuroscience 2011: 156869. Pallesen, K. J., E. Brattico, C. J. Bailey, A. Korvenoja, J. Koivisto, and A. Gjedde. 2010. ‘‘Cognitive Control in Auditory Working Memory is Enhanced in Musicians.’’ PLoS One 5: e11120. Pantev, C., and S. C. Herholz. 2011. ‘‘Plasticity of the Human Auditory Cortex Related to Musical Training.’’ Neuroscience & Biobehavioral Reviews 35: 2140–2154. Pantev, C., L. E. Roberts, M. Schulz, A. Engelien, and B. Ross. 2001. ‘‘Timbre-Specific Enhancement of Auditory Cortical Representations in Musicians.’’ NeuroReport 12: 169–174. Parbery-Clark, A., E. Skoe, and N. Kraus. 2009a. ‘‘Musical Experience Limits the Degradative Effects of Background Noise on the Neural Processing of Sound.’’ The Journal of Neuroscience 29: 14100–14107. Parbery-Clark, A., E. Skoe, C. Lam, and N. Kraus. 2009b. ‘‘Musician Enhancement for Speech-in-Noise.’’ Ear and Hearing 30: 653–661. Parbery-Clark, A., S. Anderson, E. Hittner, and N. Kraus. 2012a. ‘‘Musical Experience Offsets Age-Related Delays in Neural Timing.’’ Neurobiology of Aging 7: 1483.e1–1483.e4. Parbery-Clark, A., A. Tierney, D. L. Strait, and N. Kraus. 2012b. ‘‘Musicians have Fine-Tuned Neural Distinction of Speech Syllables.’’ Neuroscience 219: 111–119. Patel, A. D. 2011. ‘‘Why Would Musical Training Benefit the Neural Encoding of Speech? The OPERA Hypothesis. The Relationship Between Music and Language.’’ Frontiers in Psychology 2: 142. Patel, A. D., and E. Balaban. 2000. ‘‘Temporal Patterns of Human Cortical Activity Reflect Tone Sequence Structure.’’ Nature 404: 80–84. Peter, V., K. Wong, V. K. Narne, M. Sharma, S. C. Purdy, and C. McMahon. 2014. ‘‘Assessing Spectral and Temporal Processing in Children and Adults Using Temporal Modulation Transfer Function (TMTF), Iterated Ripple Noise (IRN) Perception, and Spectral Ripple Discrimination (SRD).’’ Journal of the American Academy of Audiology 25: 210–218. Rammsayer, T., and E. Altenmu¨ller. 2006. ‘‘Temporal Information Processing in Musicians and Nonmusicians.’’ Music Perception: An Interdisciplinary Journal 24: 37–48. Ruggles, D. R., R. L. Freyman, and A. J. Oxenham. 2014. ‘‘Influence of Musical Training on Understanding Voiced and Whispered Speech in Noise.’’ PLoS One 9: e86980. Schneider, P., M. Scherg, H. G. Dosch, H. J. Specht, A. Gutschalk, and A. Rupp. 2002. ‘‘Morphology of Heschl’s Gyrus Reflects Enhanced Activation in the Auditory Cortex of Musicians.’’ Nature Neuroscience 5: 688–694.

Downloaded by [Australian Catholic University] at 15:29 07 October 2017

Enhanced speech perception in noise and CAEPs in musicians Seppa¨nen, M., J. Ha¨ma¨la¨inen, A.-K. Pesonen, and M. Tervaniemi. 2012. ‘‘Music Training Enhances Rapid Neural Plasticity of N1 and P2 Source Activation for Unattended Sounds.’’ Frontiers in Human Neuroscience 6: 43. Shahin, A., D. J. Bosnyak, L. J. Trainor, and L. E. Roberts. 2003. ‘‘Enhancement of Neuroplastic P2 and N1c Auditory Evoked Potentials in Musicians.’’ The Journal of Neuroscience 23: 5545–5552. Sharma, M., S. Purdy, P. Newall, K. Wheldall, R. Beaman, and H. Dillon. 2006. ‘‘Electrophysiological and Behavioral Evidence of Auditory Processing Deficits in Children with Reading Disorder.’’ Clinical Neurophysiology 117: 1130–1144. Sharma, M., S. C. Purdy, K. J. Munro, K. Sawaya, and V. Peter. 2014. ‘‘Effects of Broadband Noise on Cortical Evoked Auditory Responses at Different Loudness Levels in Young Adults.’’ NeuroReport 25: 312–319. Swaminathan, J., C. R. Mason, T. M. Streeter, V. Best, G. Kidd, Jr, and A. D. Patel. 2015. ‘‘Musical Training, Individual Differences and the Cocktail Party Problem.’’ Scientific Reports 5: 11628. Tervaniemi, M., V. Just, S. Koelsch, A. Widmann, and E. Schro¨ger. 2005. ‘‘Pitch Discrimination Accuracy in Musicians vs Nonmusicians: An

13

Event-Related Potential and Behavioral Study.’’ Experimental Brain Research 161: 1–10. Thornton, A. R. D., M. Harmer, and B. A. Lavoie. 2007. ‘‘Selective Attention Increases the Temporal Precision of the Auditory N 100 Event-Related Potential.’’ Hearing Research 230: 73–79. Yeend, I., E. F. Beach, M. Sharma, and H. Dillon. 2017. ‘‘The Effects of Noise Exposure and Musical Training on Suprathreshjold Auditory Processing and Speech Perception in Noise.’’ Hearing Research 353: 224–236. Zendel, B. R., and C. Alain. 2009. ‘‘Concurrent Sound Segregation is Enhanced in Musicians.’’ Journal of Cognitive Neuroscience 21: 1488–1498. Zendel, B. R., and C. Alain. 2012. ‘‘Musicians Experience Less Age-Related Decline in Central Auditory Processing.’’ Psychology and Aging 27: 410–417. Zendel, B. R., C. D. Tremblay, S. Belleville, and I. Peretz. 2015. ‘‘The Impact of Musicianship on the Cortical Mechanisms Related to Separating Speech from Background Noise.’’ Journal of Cognitive Neuroscience 27: 1044–1059.

Supplementary material available online