International Journal of Pediatric Otorhinolaryngology 78 (2014) 1945–1952

Contents lists available at ScienceDirect

International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl

Mandarin lexical tones identification among children with cochlear implants or hearing aids Aifeng Li a,b,c, Ningyu Wang a,b,c,*, Jinlan Li a,b,c, Juan Zhang a,b,c, Zhiyong Liu a,b,c a

Department of Otorhinolaryngology Head and Neck Surgery, Beijing Chao-Yang Hospital, Capital Medical University, PR China College of Otolaryngology, Capital Medical University, PR China c Key Laboratory of Otolaryngology Head and Neck Surgery (Capital Medical University), Ministry of Education, Beijing 100020, PR China b

A R T I C L E I N F O

A B S T R A C T

Article history: Received 21 February 2014 Received in revised form 21 August 2014 Accepted 24 August 2014 Available online 1 September 2014

Objectives: Mandarin Chinese is a lexical tone language that has four tones, with a change in tone denoting a change in lexical meaning. There are few studies regarding lexical tone identification abilities in deafened children using either cochlear implants (CIs) or hearing aids (HAs). Furthermore, no study has compared the lexical tone identification abilities of deafened children with their hearing devices turned on and off. The present study aimed to investigate the lexical tone identification abilities of deafened children with CIs or HAs. Methods: Forty prelingually deafened children (20 with CIs and 20 with HAs) participated in the study. In the HA group, 20 children were binaurally aided. In the CI group, all of the children were unilaterally implanted. All of the subjects completed a computerized lexical tone pairs test with their hearing devices turned on and off. The correct answers of all items were recorded as the total score and the correct answers of the tone pairs were recorded as subtotal scores. Results: No significant differences in the tone pair identification scores were found between the CI group and HA group either with the devices turned on or off (t = 1.62, p = 0.11; t = 1.863, p = 0.07, respectively). The scores in the aided condition were higher than in the unaided condition regardless of the device used (t = 22.09, p < 0.001, in the HA group; t = 20.20, p < 0.001, in the CI group). Significantly higher scores were found in the tone pairs that contained tone 4. Age at fitting of the devices was correlated with tone identification abilities in both the CI and HA groups. Other demographic factors were not correlated with tone identification ability. Conclusions: The hearing device, whether a hearing aid or cochlear implant, is beneficial for tone identification. The lexical tone identification abilities were similar regardless of whether the subjects wore a HA or CI. Lexical tone pairs with different durations and dissimilar tone contour patterns are more easily identified. Receiving devices at earlier age tends to produce better lexical tone identification abilities in prelingually deafened children. ß 2014 Elsevier Ireland Ltd. All rights reserved.

Keywords: Cochlear implant Hearing aid Tone identification Mandarin lexical tone Pediatric

1. Introduction According to Han’s statistics, nearly four million people have profound hearing loss (PTA 0.5–2 kHz > 90 dB as in the WHO classification) in China, a country of 1.322 billion people [1]. Prior to November 2012, a very small proportion of children with profound hearing losses in China, approximately 15,000, had

* Corresponding author at: Department of Otorhinolaryngology Head and Neck Surgery, Beijing Chao-Yang Hospital, Capital Medical University, Beijing, PR China. Tel.: +861085231348. E-mail addresses: [email protected], [email protected] (N. Wang). http://dx.doi.org/10.1016/j.ijporl.2014.08.033 0165-5876/ß 2014 Elsevier Ireland Ltd. All rights reserved.

received cochlear implants [2]. Compared to the 5000 cases in January 2008, the number of children with a prelingual profound hearing loss who receive cochlear implants is now increasing rapidly. Therefore, it is necessary to help more hearing-impaired children receive hearing devices (e.g., cochlear implant or hearing aid) in order to achieve a high level of hearing ability. Western languages, such as English, French, German and Dutch, are non-tonal Indo-European languages. In contrast, Chinese (which is spoken by one-quarter of all people in the world), including Mandarin, Cantonese, and other dialects, is a tonal SinoTibetan language. Mandarin, China’s official language, is the language selected for this study. Mandarin Chinese syllables are produced with one of four lexical tones: tone 1 (high-flat), tone 2 (rising), tone 3 (falling–rising), and tone 4 (falling). The same

1946

A. Li et al. / International Journal of Pediatric Otorhinolaryngology 78 (2014) 1945–1952

syllable with different tones can convey vastly different meanings [3]. Hence, it is important to identify and produce tones correctly. Past studies have shown that having all parts of the basement membrane actively stimulated and an increased insertion depth can help CI users to improve their speech identification abilities. However, the dominant difference between the four Mandarin tones is the variation in fundamental frequency (F0). A few studies have shown that fine structure information and temporal envelope information were the two underlying perceptual factors by which Mandarin tones are identified. Fine structure information refers to the fundamental frequency and its harmonics. Temporal envelope information refers to the amplitude envelope and duration [4]. In the available CI strategies, few channels are allocated to representing the full spectrum of F0, which has been suggested to provide important acoustic cues to differentiate between lexical tones. A lack of a full presentation of F0, a mismatched cutoff frequency, fewer total channels, and a loss of the fine structure may contribute to relatively poor tonal perception. The four tones in Mandarin produce a total of six tone pairs. Children with profound hearing loss have difficulty identifying the tone system; it affects the ability to produce correct tones as well as mental development. Therefore, the identification of the six tone pairs could be considered as a method for improving acoustic sensitivity in hearing-impaired children in rehabilitation programs. There are comparatively few CI studies focused on the contribution of tone identification to understand Mandarin. Peng [5] examined all six tone pairs with 30 CI users from six to twelve years of age and found that children more easily identified tone pairs that contained tone 4 than pairs that did not contain this tone. Liu [6] tested the Mandarin tone identification abilities of six children with prelingual, profound hearing losses who used CIs. The results were consistent with those found by Peng. Tone 4 had higher identification scores than other tones. That is, words pronounced with tone 4 were more easily identified than words pronounced with other tones. As we know, the basic operating principles of hearing aids and cochlear implants are different. Conventional hearing aids amplify sound signals that are then perceived by users. Cochlear implants, in contrast, bypass damaged hair cells and directly stimulate the auditory nerve by converting mechanical sound energy into an electrical stimulus [7]. Many studies have compared the speech and language skills of English-speaking children using CIs to children fitted with conventional hearing aids [8–10] and found that CI devices can offer better speech and language performance than conventional hearing aids. Tomblin [11] examined English language achievement in 29 children with profound hearing loss using CIs and 29 deaf children using HAs. It was found that CI users achieved significantly better scores than HA users with respect to knowledge of English grammar. Several studies from Hong Kong have investigated Cantonese speech perception with CIs and HAs and obtained different results. Lee [7] reported results from 18 HA users and 34 CI users who identified 15 Cantonese tone pairs. There was no significant difference in tone identification between children with HAs and CIs in the aided condition. Law [12] examined the phonological abilities of children with hearing loss fitted with HAs and CIs. The results showed that Cantonese children with CIs had better phonological skills than children with HAs when they had similar degrees of hearing loss. The aforementioned tests have shown that there are no consistent views about improvements in tone identification in hearing aid and cochlear implant users. Although studies reported that the tone identification abilities of cochlear implant users were better than those of hearing aid users, both devices are insufficient in improving tone identification in children with profound hearing loss. Liu [2] investigated the Mandarin consonant identification abilities of children with profound hearing loss, 41 with CIs and 26

with HAs. The results showed that there was no significant difference in consonant identification scores between the CI and HA users. Thus, it is essential to know which device (hearing aid or cochlear implant) is best for the perception of tones in Mandarin. All current studies compared tone perception between CIs and HAs with them turned on. To date, most pediatric CI studies have either compared speech-understanding abilities before and after cochlear implantation [13], or they compared the performances of subjects using CIs to normal hearing subjects [14]. Similar research comparing the unaided to the aided condition among children with CIs and HAs is lacking; past studies showed that hearing aids can benefit moderate to severe, but not profound, hearing losses. Cochlear implants can be beneficial for profound hearing losses. The extent to which CIs and HAs can assist with tone identification remains unclear. The first aim of the present study was to investigate the function of hearing devices in the identification of Mandarin lexical tone pairs when the tones were presented at subjects’ most comfortable loudness levels. Tone pair identification performance was compared with the hearing devices turned on and off. The second aim was to determine which Mandarin tone pairs the subjects identified most accurately. Finally, a multiple linear regression analysis was performed to see whether demographic factors (e.g., age at implantation, age at testing, duration of device usage, degree of hearing loss) predicted speech performance. 2. Methods and materials 2.1. Subjects Forty school-age Mandarin-speaking children (22 male and 18 female) with prelingual, profound hearing losses (PTA > 90 dB) were divided into two groups based on their primary type of hearing device, that is, a hearing aid group (HA, n = 20) and a cochlear implant group (CI, n = 20). In the HA group, 20 children were binaurally aided. Their mean age was 9 years, 10 months (9;10), with a range of 7;9–12;6. The mean length of their hearing aid experience was 4;7, ranging from 3;7 to 5;8. The mean age at receiving hearing aids was 5;3, with a range of 3;6–7;6. The children with HAs used the better hearing ear as the test side. In the CI group, the children had received a fully inserted unilateral cochlear implant that had been in place for at least 44 months at the time of data collection. Their mean age was 8;7, with a range of 6;0–11;1. The mean length of their cochlear implant experience was 4;5, with a range of 3;8–6;7. The mean age at implantation was 4;1, with a range of 2;0–6;8. Five subjects were implanted in the left ear, and 15 subjects were implanted in the right ear. Four implanted children also used one hearing aid on the nonimplanted ear. Detailed demographic information is summarized in Table 1. Subject selection criteria included having a bilateral profound sensorineural hearing loss and no evidence of mental retardation, learning disorders or behavior disorders. Conversely, none of the children were in a gifted or talented educational program. Mandarin Chinese was their native language. All of the children were in educational and home environments that provided them with total oral communication and used hearing devices that had been prescribed by qualified audiologists. The types and models of the devices were not controlled. The use of human subjects was reviewed and approved by the ethics committee of Beijing Chao-yang Hospital. 2.2. Materials A two-alternative forced-choice tone contrast test was used. The subject’s task was to discriminate between a pair of monosyllabic words in which only the tone patterns differed

A. Li et al. / International Journal of Pediatric Otorhinolaryngology 78 (2014) 1945–1952

1947

Table 1 Demographic characteristics of the children.

Age at testing Unaided PTAa in the left ear Unaided PTAa in the right ear Age at receiving current device Hearing age Brand of CI device Speech-coding strategy a

Hearing aid group

Cochlear implant group

N = 20

N = 20

118 months (range, 93–150) 101 dB 104 dB 63 months (range, 42–90) 55 months (range, 43–68)

103 months (range, 72–133) 112 dB 113 dB 49 months (range, 24–80) 53 months (range, 44–79) Nucleus-24 Advance combination encoder

PTA (in dB HL) was taken for the two ears at the frequencies of 500, 1000, 2000, and 4000 Hz before the surgery.

[e.g., chuang (tone 1), ‘‘a window,’’ and chuang (tone 2), ‘‘a bed’’]. The four Mandarin tones produced six possible tone pairs (i.e., tone 1 versus tone 2, tone 1 versus tone 3, tone 1 versus tone 4, tone 2 versus tone 3, tone 2 versus tone 4, and tone 3 versus tone 4). We provided three sets of words to allow the subjects to choose one set that they knew best. The chosen set was then used in the subsequent test for that subject. For each of the words, a picture (along with the Chinese characters and phonemic spelling) representing the meaning of the word was presented visually. We chose monosyllabic words based on the vocabulary level of young children. The two words in each test item were as follows: (1) words familiar to the child, (2) monosyllabic words, (3) minimal pairs contrasting in tone only, and (4) suitable for representation in picture format. Therefore, while the contents of the test materials may have differed, the test batteries met the basic requirements for developing speech testing materials: familiarity, homogeneity, phonemic balance, and equivalence [15]. The stimuli were recorded in a double-walled sound-treated booth (Acoustic Systems). An Electro-Voice omnidirectional microphone connected to an external sound card was used for the recording. The distance between the speaker’s lips and the microphone was kept at approximately 10 cm. Syllables of each pair of tones were recorded multiple times. The speech materials were recorded by a female, native Mandarin-speaking adult with normal hearing whose average vocal F0 was 259 Hz. A 44.1 kHz sampling rate and 16-bit resolution were used.

devices turned on (aided condition). The order of the oral presentation of different target tones was randomly arranged to prevent any memory effects. The tasks were performed in the unaided condition before being performed in the aided condition to minimize any possible learning effects because the aided condition would allow the child to perceive more information than the unaided condition. 3. Result Rating values from the test administrator represented the degree of correctness. Each of the values was converted to a percent value that represented the degree of accuracy, where 0 represented 0% and 4 represented 100%. A one-sample Kolmogorov–Smirnov test was used to assess the normality of the dataset, and the results showed that all of the subtotal scores and total scores were normally distributed. Fig. 1 plots the total scores of the two groups for the identification of tone pairs. The mean unaided scores in the HA and CI groups were 18% and 15%, respectively. The mean aided scores in the HA and CI groups were significantly higher (66%, ranging from 54% to 79% and 72%, ranging from 54% to 83%, respectively), which was

2.3. Procedures The tone identification test was conducted in a sound-treated room. A graphical user interface displayed two pictures representing a pair of tones on a laptop computer screen. The stimuli were presented through a loudspeaker at 65 dB(A) as measured at the ear level of the subjects. The speech stimulus corresponding to one of the pictures was presented through a pair of loudspeakers located in front of the subjects. The subjects were required to point to the picture corresponding to the word they heard. A pilot study of 20 normal-hearing, native Mandarin-speaking children of three to nine years of age confirmed the applicability of the words. The average accuracy of overall tone identification of the pairs was 98% for the normal hearing subjects. The tone identification performance for each tone showed little difference among the four tones for the normal-hearing subjects. Before the formal testing, all children practiced for a few minutes to familiarize themselves with the task and to ensure they understood the procedures. Four trials were performed for each tone pair, and each child’s response was entered on site by the test administrator. The child received one point if she or he identified a target tone accurately. The proportion of total correct responses was computed as the total average score, while the proportion of correct responses for each tone pair was computed as the subtotal score of the tone pair. All subjects performed the test twice, once with their hearing devices turned off (unaided condition), and the other time with the

Fig. 1. Box plot illustrates accuracy (in percentage) of tone pairs perception scores between the hearing aid and cochlear implant groups with the aided and unaided situation; y-axis displays total scores of tone pairs perception. Median is displayed by the line across each gray box; upper and lower bounds of each gray box represent quartiles.

1948

A. Li et al. / International Journal of Pediatric Otorhinolaryngology 78 (2014) 1945–1952

Fig. 2. Group mean data on tone pairs identification with the devices turned on. Each bar represents data from one of the two hearing devices: cochlear implants (gray bar) and hearing aid (slash bar). The x-axis represents the six tone pairs; y-axis represents identification scores of each tone pair. The error bar represents SD.

significantly higher than the 50% chance level (t (19) = 9.53, p < 0.001; t (19) = 8.97, p < 0.001, respectively). Fig. 1 shows that a significant difference exists between the aided and the unaided tone perception scores in the HA and CI groups (t = 22.09, p < 0.001; t = 20.20, p < 0.001, respectively). There is no significant difference between the CI and HA groups in either the unaided or the aided condition (t = 1.62, p = 0.11; t = 1.863, p = 0.07, respectively). The subtotal scores in the unaided and aided conditions in the two groups were compared separately. In both groups, the increases in scores in the aided conditions were all significant across the six tone pairs, with p < 0.001. With regard to the identification of specific tone pairs, the aided scores of the two groups were used. Fig. 2 shows a bar chart of the children’s identification scores for the six tone pairs in the two groups. A randomized block design ANOVA was used to test whether performances differed in the two groups and six tone pairs. The results revealed that the main effect of the tone pairs was significant (F = 5.54, p < 0.001), as was the main effect of the subjects (F = 2.53, p < 0.001). The interaction between the subjects and tone pairs was not significant (p > 0.05). Post-hoc Bonferroni pair-wise comparisons showed that tone pair identification was better in subjects with CIs than HAs (adjusted p < 0.01). The differences in scores for the following six tone pairs were significant: between tone 1 versus tone 4 and tone 1 versus tone 3 (p = 0.002), tone 1 versus tone 2 and tone 1 versus tone 4 (p = 0.008), tone 1 versus tone 4 and tone 3 versus tone 2 (p = 0.015), tone 2 versus tone 4 and tone 2 versus tone 3 (p = 0.004), and tone 3 versus tone 4 and tone 2 versus tone 3 (p = 0.008). There were no significant differences between the remaining tone pairs. To assess which demographic factors may have contributed to performance, single and multiple linear regression analyses were performed. Fig. 3 shows the relationship between the age at receiving devices and mean tone pair identification scores in the CI and HA groups (r = 0.82, p < 0.001; r = 0.92, p < 0.001, respectively). Fig. 4 shows the duration of receiving a hearing device was not correlated with tone identification (r = 0.12, p = 0.96 in the CI group; r = 0.05, p = 0.84 in the HA group). As shown in Fig. 5, the

Fig. 3. Scatter plot illustrates distribution of accuracy (in percentage) of tone pairs identification (y-axis) as a function of age at receiving devices (x-axis). Children with the cochlear implants are marked with an asterisk; those with the hearing aids are marked with a circle.

correlation between tone pair identification and chronological age was significant (r = 0.70, p < 0.001 in the CI group; r = 0.81, p < 0.001 in the HA group). The significant correlation between tone pair identification and chronological age was a result of a high

Fig. 4. Scatter plot illustrates distribution of accuracy (in percentage) of tone pairs identification (y-axis) as a function of duration of time using devices (x-axis). Children with the cochlear implants are marked with a asterisk; those with the hearing aids are marked with a circle.

A. Li et al. / International Journal of Pediatric Otorhinolaryngology 78 (2014) 1945–1952

1949

Fig. 5. Scatter plot illustrates distribution of accuracy (in percentage) of tone pairs recognition (Y-axis) as a function of test age (X-axis). Children with the cochlear implants are marked with an asterisk; those with the hearing aids are marked with a circle.

correlation between chronological age and age at receiving devices (r = 0.88; p < 0.001). Fig. 6 shows that no correlation was found between tone pair identification and the average degree of hearing loss in the CI group and HA group (r = 0.24, p = 0.30; r = 0.29, p = 0.20, respectively). Age at receiving devices was the only significant factor in the present study. No significant correlations were found between any of other variables and the scores. 4. Discussion

Fig. 6. Scatter plot illustrates distribution of accuracy (in percentage) of tone pairs recognition (y-axis) as a function of degree of hearing loss (x-axis). Children with the cochlear implants are marked with an asterisk; those with the hearing aids are marked with a circle.

In general, hearing devices, either hearing aids or cochlear implants, significantly improved the tone perception performance of Mandarin-speaking pediatric subjects. No major difference in tone perception performance was found between the HA group and CI group, which indicated that the general functional gain in Mandarin tone perception was similar for these hearing devices. The data showed that the differences in the identification scores for the six tone pairs when the hearing devices were turned on were statistically significant. Age at receiving devices was correlated with tone identification performance in the two groups. Below we discuss the results in detail. Past studies have compared the performances of hearingimpaired children with their hearing devices turned on and turned off. For example, Lee [16] found that hearing aid use is beneficial for Cantonese tone perception in children (with a mean age of five years) with moderate to severe hearing loss. When hearing losses are greater than 90 dB, i.e., in children who are classified as having profound hearing losses, hearing aids are not effective in aiding Cantonese tone perception. Contrary to the past study, the results of the present study show that hearing devices, either cochlear

1950

A. Li et al. / International Journal of Pediatric Otorhinolaryngology 78 (2014) 1945–1952

implants or hearing aids, under the aided condition significantly improved the tone perception performance of Mandarin-speaking children with prelingual, profound hearing impairments. There are two possible explanations for the divide in tone perception abilities of hearing-impaired children. First, different study designs and larger sample sizes may contribute to the result obtained. Only three tone pairs were tested in Lee’s study and the number of subjects was small, at only eight children. In the present study, 40 children with hearing loss were included and all six tone pairs were tested. The larger sample size and full-scale test items may make the result more accurate. Second, the higher aided tone perception scores may also be due to the older age of the subjects tested in the present study (118 months on average). The difference in tone perception scores between the CI and HA groups was not significant. The aided tone perception score was 72% for the CI group and 66% for the HA group. The results show that the ability to perceive Mandarin tones is similar for both cochlear implants and hearing aids, which is different from the well-established evidence. Previous studies indicated that Englishspeaking CI users had better tone perception performance than HA users [8,17,10,18–20,11]. Geers [21] compared the language development of a group of 13 children who received CIs to a similar group of children fitted with HAs or tactile aids over three years. All three groups were provided with intensive oral speech and language training. The language growth of children with CIs equaled or exceeded the growth of the other groups in receptive and expressive measures of spoken English. In particular, Miyamoto [22] examined scores on the Reynell Developmental Language Scales – Revised over the first year of CI experience and compared these changes to changes in those who had not received CIs but had similar demographic data. They found that even 12 months of CI experience was sufficient to produce significant gains in expressive language over predictions based on non-implanted children. The tone perception performance seen in this study was also comparable to that of previous studies with Mandarin-speaking pediatric subjects. Zhou [2] measured Mandarin consonant pair identification performance among children with profound hearing loss, 41 who used CIs and 26 who used HAs, and 30 children with normal hearing. The study showed no significant difference between the CI and HA groups. Consistent with Zhou’s result, Lee [7] evaluated Cantonese lexical tone pair identification between 18 children with profound hearing loss using HAs and 34 using CIs and found no significant difference between the two groups. The results of the present study are not consistent with the well-established evidence of CI users having better tone perception than HA users. The following reasons may explain this result. Comparisons between CI and HA groups are often difficult because of differences in demographic factors such as age at testing, length of device use, degree of hearing loss, age at implantation, age at receiving a hearing aid, and educational setting [2]. In the present study design, the two groups of subjects were not matched according to their amount of hearing loss, hearing experience, or rehabilitation setting. Factors such as chronological age and maturity of cognitive function are likely to interact with hearing age. Further, it is possible that some intrinsic traits may contribute to the current results. Correlation and regression analyses were used to examine the relationship between tone pair identification scores and hearing history variables in the aided condition. The result showed that there was no significant relationship between tone identification scores and the examined factors, such as degree of hearing loss and duration of device usage. Age at testing and age at receiving hearing devices were found to be significantly correlated with tone identification scores (see the following section for a discussion of age at implantation). Moreover, the covariance and regression

were not used to analyze the large number of potential factors and sample capacity. Further research with a more detailed study design will be necessary to determine whether these two groups will diverge in tone pair identification abilities. There is no significant difference in tone pair identification scores between the CI group and HA group under the aided condition. The mean degree of hearing loss in the CI group was 113 dB, and 80% of these children had profound hearing impairment (hearing loss of 100 dB or more), while the mean degree of hearing loss in the HA group was 103 dB, and 65% of these children had profound hearing impairment. It seems that, among children with the same tone identification scores, the degree of hearing loss in the HA group was lower than that of the CI group. The results showed that among children with hearing losses over 100 dB, cochlear implants are better than hearing aids for tone pair identification. The effect of cochlear implants on tone identification has been well documented. Iwasaki [23] reported that speech discrimination scores and intelligibility ratings were higher in CI users than in HA users when the degree of hearing loss in CIs was higher than in HAs. Sininger [24] tested 44 infants and toddlers who were identified with mild to profound bilateral hearing losses and the results demonstrated a significant relationship between the degree of hearing loss and language outcomes. Degree of hearing loss was an important factor in the modeling of speech production and spoken language outcomes. In the present study, the identification scores of the six tone pairs under the aided condition were different. The average score was highest for tone 2 versus tone 4 and lowest for tone 1 versus tone 3. The other tone pairs children identified, from easiest to most difficult, were as follows: tone 1 versus tone 4, tone 3 versus tone 4, tone 1 versus tone 2, and tone 2 versus tone 3. This tone pair identification performance was comparable to that of previous studies with Mandarin-speaking pediatric subjects. Peng [5] studied tone identification in 30 children with profound hearing loss using CIs; the mean score was 72.9% (chance level = 50%). The identification scores were significantly higher for pairs that contained tone 4 compared to pairs without that tone. In Xu’s test, tone identification in normal-hearing listeners was reduced by 15% points by using vocoder-processed tone tokens and removing the duration cues [25]. The study showed that removal of the duration cue affected the performance of tone 4 the most. That is, when the normal duration exists, tone 4 can be identified more easily than others. The intrinsic features of tone 4 may help to explain these results. Tone 4 (the high-falling tone) has the shortest duration of the tones. It is likely that the children with CIs were able to use duration cues to distinguish tone 4 from other tones. In contrast, other tones, lacking such a patent duration cue, were not identified easily when they were paired. The results are consistent with those found in a study by Whalen [26]. Moreover, patterns of Mandarin tones are defined by pitch contours and the height of the speech signals. Whalen [26] noted that Mandarin tone information might be carried by the amplitude contour (temporal envelope) of the speech signal. They argued that the relatively significant amplitude change was important in perceptions of Mandarin tones. In Mandarin tones, tone 4 is identified more easily because of the greater difference in its amplitude contour compared to the other tones. We infer that the greater the difference in amplitude contours between tone pairs (i.e., tone 4 versus tone 1, tone 4 versus tone 2, and tone 4 versus tone 3), the easier it is for subjects to correctly identify them. This can be supported by the present findings in which the subjects can identify tone pairs containing tone 4, better than other tone pairs not containing tone 4. The relative difficulty in identifying other tone pairs may be attributed to the smaller differences in amplitude contours. The results are consistent with those found in a study by Mannell [27]. Mannell measured contributions of the

A. Li et al. / International Journal of Pediatric Otorhinolaryngology 78 (2014) 1945–1952

temporal envelope to lexical tone identification in 31 subjects with different levels of hearing loss. A significant negative correlation was discovered between tone identification performance and audiometric hearing thresholds. In summary, amplitude contour and tone duration may be two of the contributors to Mandarin tone identification for hearing-impaired subjects in the present study. Past studies have predicted that age at receiving a hearing device had a positive effect on speech performance [28]. Shea [29] reported that age at receiving devices was positively correlated with speech performance in children with profound hearing losses who used CIs. Harrison [30] reported that children implanted at an early age outperformed those implanted at a later age, even after long-term CI experience. The effect of receiving hearing devices early on identification performance has been well documented in many western languages [31–33] and in Mandarin [1]. The results of this study are in accord with past studies and show progressively stronger identification performance. Although the mean age at receiving devices was four years in this study, beyond the critical period for auditory development, tone identification performance was not affected. Sharma [28] examined the consequences of different ages of implantation on the development of the human central auditory system and found that development and plasticity of the central auditory system continue in some, but not all, children until approximately age seven. After age seven, plasticity is greatly reduced. I believe that in my experiment, this was part of the reason for the results. It is necessary to note that there is a correlation between tone identification and chronological age. The significant correlation between tone identification and chronological age was most likely a result of a high correlation between chronological age and age at receiving devices in our subjects (r = 0.88; p < 0.001). According to the present results, factors such as the degree of hearing loss and duration of time using the devices were not related to the tone identification performance (p > 0.05). Some substantial characteristics of the subjects were not assessed in the present study, such as learning effects and ceiling effects, which had effects on tone identification performance. Therefore, the function of one factor may be amplified or reduced when other factors were ignored. 5. Conclusions The present study revealed several significant findings regarding tone identification performance in Mandarin-speaking children with profound hearing losses: (1) In sum, both hearing devices were beneficial for enhancing tone identification performance in hearing-impaired children. However, the benefit from cochlear implants was greater than that of hearing aids among the children with hearing threshold above 100 dB. (2) The aided identification scores of the specific tone pairs were different among the six contrast pairs. Tone 2 versus tone 4, tone 1 versus tone 4, and tone 3 versus tone 4 had relatively higher scores compared to the other tone pairs, suggesting that the identification performance of the four Mandarin tones does not develop in parallel. Amplitude contours and tone durations may contribute to why certain tones or tone pairs were much easier than others to identify accurately. (3) Age at implantation strongly affected the tone identification performance of the hearing-impaired children. Other factors, such as the degree of hearing loss, duration of time using devices, and age at testing, were not well predicted by the tone identification performance.

1951

Acknowledgments This research was supported by the Capital Health Research and Development of Special, China (No. 2011200302). The authors thanked the children and parents who participated in the study. We also thanked Wei Wei for the help in collecting and analyzing data. References [1] D. Han, B. Liu, N. Zhou, X. Chen, Y. Kong, H. Liu, Lexical tone perception with HiResolution and HiResolution 120 sound-processing strategies in pediatric Mandarin-speaking cochlear implant users, Ear Hear. 2 (2009) 169–177. [2] Q. Liu, N. Zhou, R. Berger, D. Huang, L. Xu, Mandarin consonant contrast recognition among children with cochlear implants or hearing aids and normal-hearing children, Otol. Neurotol. 3 (2013) 471–476. [3] S. Prom-on, Y. Xu, B. Thipakorn, Modeling tone and intonation in Mandarin and English as a process of target approximation, J. Acoust. Soc. Am. 1 (2009) 405–424. [4] Y.Y. Kong, F.G. Zeng, Temporal and spectral cues in Mandarin tone recognition, J. Acoust. Soc. Am. 5 (Pt 1) (2006) 2830–2840. [5] S.C. Peng, J.B. Tomblin, H. Cheung, Y.S. Lin, L.S. Wang, Perception and production of mandarin tones in prelingually deaf children with cochlear implants, Ear Hear. 3 (2004) 251–264. [6] T.C. Liu, H.P. Chen, H.C. Lin, Effects of limiting the number of active electrodes on Mandarin tone perception in young children using cochlear implants, Acta Otolaryngol. 10 (2004) 1149–1154. [7] K.Y. Lee, C.A. van Hasselt, M.C. Tong, Lexical tone perception ability of profoundly hearing-impaired children: performance of cochlear implant and hearing aid users, Otol. Neurotol. 7 (2010) 1079–1087. [8] C.M. Connor, H.K. Craig, S.W. Raudenbush, K. Heavner, T.A. Zwolan, The age at which young deaf children receive cochlear implants and their vocabulary and speech-production growth: is there an added value for early implantation? Ear Hear. 6 (2006) 628–644. [9] K.M. Ceh, D.M. Bervinchak, H.W. Francis, Early literacy gains in children with cochlear implants, Otol. Neurotol. 3 (2013) 416–421. [10] M.J. Osberger, A.M. Robbins, R.T. Miyamoto, S.W. Berry, W.A. Myres, K.S. Kessler, et al., Speech perception abilities of children with cochlear implants, tactile aids, or hearing aids, Am. J. Otol. 12 (1991) 105–115. [11] J.B. Tomblin, L. Spencer, S. Flock, R. Tyler, B. Gantz, A comparison of language achievement in children with cochlear implants and children using hearing aids, J. Speech Lang. Hear. Res. 2 (1999) 497–509. [12] Z.W. Law, L.K. So, Phonological abilities of hearing-impaired Cantonese-speaking children with cochlear implants or hearing AIDS, J. Speech Lang. Hear. Res. 6 (2006) 1342–1353. [13] J.L. Wu, H.M. Yang, Speech perception of Mandarin Chinese speaking young children after cochlear implant use: effect of age at implantation, Int. J. Pediatr. Otorhinolaryngol. 3 (2003) 247–253. [14] S. Wang, R. Mannell, P. Newall, D. Han, Contribution of spectral cues to mandarin lexical tone recognition in normal-hearing and hearing-impaired Mandarin Chinese speakers, Ear Hear. 1 (2011) 97–103. [15] K.S. Tsai, L.H. Tseng, C.J. Wu, S.T. Young, Development of a mandarin monosyllable recognition test, Ear Hear. 1 (2009) 90–99. [16] K.Y. Lee, C.A. van Hasselt, M.C. Tong, Tone perception in Cantonese-speaking children with hearing aids, Ann. Otol. Rhinol. Laryngol. 4 (2008) 313–316. [17] A. Scorpecci, F. Zagari, G. Mari, S. Giannantonio, L. D’Alatri, W. Di Nardo, G. Paludetti, Investigation on the music perception skills of Italian children with cochlear implants, Int. J. Pediatr. Otorhinolaryngol. 10 (2012) 1507–1514. [18] P.W. Dawson, P.J. Blamey, L.C. Rowland, S.J. Dettman, G.M. Clark, P.A. Busby, Cochlear implants in children, adolescents, and prelinguistically deafened adults: speech perception, J. Speech Hear. Res. 2 (1992) 401–417. [19] J.S. Moog, A.E. Geers, Assessing the benefits of cochlear implants in an oral education program, in: A.S. Uziel, M. Mondain (Eds.), Cochlear Implants in Children, 1995, 119–124. [20] A.F. Snik, A.M. Vermeulen, J.P. Brokx, C. Beijk, P. van den Broek, Speech perception performance of children with a cochlear implant compared to that of children with conventional hearing aids. I. The ‘‘equivalent hearing loss’’ concept, Acta Otolaryngol. 5 (1997) 750–754. [21] A.E. Geers, L.S. Davidson, R.M. Uchanski, J.G. Nicholas, Interdependence of linguistic and indexical speech perception skills in school-age children with early cochlear implantation, Ear Hear. 5 (2013) 562–574. [22] R.T. Miyamoto, M.A. Svirsky, A.M. Robbins, Enhancement of expressive language in prelingually deaf children with cochlear implants, Acta Otolaryngol. 2 (1997) 154– 157. [23] S. Iwasaki, S. Nishio, H. Moteki, Y. Takumi, K. Fukushima, N. Kasai, Language development in Japanese children who receive cochlear implant and/or hearing aid, Int. J. Pediatr. Otorhinolaryngol. 3 (2012) 433–438. [24] Y.S. Sininger, A. Grimes, E. Christensen, Auditory development in early amplified children: factors influencing auditory-based communication outcomes in children with hearing loss, Ear Hear. 2 (2010) 166–185. [25] L. Xu, Y. Tsai, B.E. Pfingst, Features of stimulation affecting tonal-speech perception: implications for cochlear prostheses, J. Acoust. Soc. Am. 1 (2002) 247–258. [26] D.H. Whalen, Y. Xu, Information for Mandarin tones in the amplitude contour and in brief segments, Phonetica 1 (1992) 25–47.

1952

A. Li et al. / International Journal of Pediatric Otorhinolaryngology 78 (2014) 1945–1952

[27] S. Wang, L. Xu, R. Mannell, Relative contributions of temporal envelope and fine structure cues to lexical tone recognition in hearing-impaired listeners, J. Assoc. Res. Otolaryngol. 6 (2011) 783–794. [28] A. Sharma, M.F. Dorman, A.J. Spahr, A sensitive period for the development of the central auditory system in children with cochlear implants: implications for age of implantation, Ear Hear. 6 (2002) 532–539. [29] N. Eze, E. Ofo, D. Jiang, A.F. O’Connor, Systematic review of cochlear implantation in children with developmental disability, Otol. Neurotol. 8 (2013) 1385–1393. [30] R.V. Harrison, K.A. Gordon, R.J. Mount, Is there a critical period for cochlear implantation in congenitally deaf children? Analyses of hearing and

speech perception performance after implantation, Dev. Psychobiol. 3 (2005) 252–261. [31] P.J. Govaerts, C. De Beukelaer, K. Daemers, G. De Ceulaer, M. Yperman, T. Somers, Outcome of cochlear implantation at different ages from 0 to 6 years, Otol. Neurotol. 6 (2002) 885–890. [32] A.S. Uziel, M. Sillon, A. Vieu, F. Artieres, J.P. Piron, J.P. Daures, Ten-year follow-up of a consecutive series of children with multichannel cochlear implants, Otol. Neurotol. 5 (2007) 615–628. [33] C. O’Neill, G.M. O’Donoghue, S.M. Archbold, T.P. Nikolopoulos, T. Sach, Variations in gains in auditory performance from pediatric cochlear implantation, Otol. Neurotol. 1 (2002) 44–48.

Mandarin lexical tones identification among children with cochlear implants or hearing aids.

Mandarin Chinese is a lexical tone language that has four tones, with a change in tone denoting a change in lexical meaning. There are few studies reg...
714KB Sizes 1 Downloads 8 Views