Warning: mkdir(): Permission denied in /home/virtual/lib/view_data.php on line 87 Warning: chmod() expects exactly 2 parameters, 3 given in /home/virtual/lib/view_data.php on line 88 Warning: fopen(/home/virtual/audiology/journal/upload/ip_log/ip_log_2024-10.txt): failed to open stream: No such file or directory in /home/virtual/lib/view_data.php on line 95 Warning: fwrite() expects parameter 1 to be resource, boolean given in /home/virtual/lib/view_data.php on line 96 Comparison of Speech Perception Performance According to Prosody Change Between People With Normal Hearing and Cochlear Implant Users

Comparison of Speech Perception Performance According to Prosody Change Between People With Normal Hearing and Cochlear Implant Users

Article information

J Audiol Otol. 2024;28(2):119-125
Publication date (electronic) : 2023 December 7
doi : https://doi.org/10.7874/jao.2023.00234
1Department of Speech Language Pathology, Graduate School of Interdisciplinary Therapy, Myongji University, Seoul, Korea
2Department of Communication Disorders, Ewha Womans University, Seoul, Korea
Address for correspondence Hye Yoon Seol, AuD Department of Communication Disorders, Ewha Womans University, 52 Ewhayeodae-gil, Seodaemun-gu, Seoul 03760, Korea Tel +82-2-3277-3769 E-mail seol.helena@ewha.ac.kr
Received 2023 June 22; Revised 2023 July 13; Accepted 2023 July 27.

Abstract

Background and Objectives

Cochlear implants (CIs) are well known to improve audibility and speech recognition in individuals with hearing loss, but some individuals still struggle with many aspects in communication, such as prosody. This study explores how prosodic elements are perceived by those with normal hearing (NH) and CIs.

Subjects and Methods

Thirteen individuals with NH and thirteen CI users participated in this study and completed speech perception, speech prosody perception, speech prosody production, pitch difference discrimination, and melodic contour perception testing.

Results

NH listeners performed significantly better than CI users on speech perception, speech prosody perception (except for words with neutral meaning and a negative prosody change and when words were repeated twice), pitch difference discrimination, and melodic contour perception testing. No statistical significance was observed for speech prosody production for both groups.

Conclusions

Compared to NH listeners, CI users had limited ability to recognize prosodic elements. The study findings highlight the necessity of an assessment tool and signal processing algorithm for CIs, specifically targeting prosodic elements in clinical settings.

Introduction

The speech chain represents the overall process of perception and production of speech [1]. The speech chain involves three stages: linguistic, production, and acoustic. The linguistic stage begins with a speaker encoding a message. At the production stage, speech is produced through articulation and lastly, at the acoustic stage, sound is heard by the listener through his or her auditory system. The smallest unit of sound that distinguishes meaning is the phoneme and it is a segmental element of speech. Suprasegmental or prosodic elements, on the other hand, include pitch, loudness, and length. These elements are the non-linguistic aspect of communication that conveys the emotions and intentions of the speaker, going beyond mere words. For example, when individuals express emotions, such as fear or anger, variations in stress and intonation patterns can be observed [2-4]. These suprasegmental elements can also contribute to differences in meanings and when this occurs, it is called prosodeme [5].

Cochlear implant (CI) is one of the hearing devices that can be used for individuals with severe to profound sensorineural hearing loss. A CI consists of two components: internal and external. The internal component refers to the electrode array which sends signal to the auditory system. The external component plays a key role in transferring sound from outside to the electrode array through a coil. CIs are typically recommended for those whose hearing loss cannot be managed with hearing aids [6-9]. Significant advancement in technology, surgical technique, and aesthetics has been made and the benefit of CI in speech understanding has been well noted in literature [10-18]. While CI can increase individuals’ ability to understand speech, it is important to also note that not everyone experiences this benefit; even if hearing can be restored through CIs, some individuals still struggle with communication, especially in the aspect of prosody [19-23]. Marx, et al. [19] conducted a study to investigate the discrimination performance in 21 CI users using with a statement-question intonation task and a F0 difference limen test. Here, the participants were divided into two groups (with and without residual hearing). Overall, CI users with residual hearing perceived prosody and F0 better than those without residual hearing. In Klieve and Jeanes [24], children with CIs performed below 80% on the Perception of Prosody Assessment Tool sub-tests that examined various aspects of prosody, such as stress and tone and affect, and in Lenden and Flipsen [25], the authors analyzed 40 speech samples from children with CIs and reported noticeable problems with stress and resonance quality (i.e., loudness, rate, and laryngeal quality).

A possible factor contributing to this issue is characteristics of current measurement tools in clinic. Various tests, such as speech-in-noise test, are used to assess individuals’ speech performance before and after CI surgery. However, these types of tests only focus on the segmental aspect of speech. For example, the Korean version of the Hearing in Noise Test is commonly utilized to assess individuals’ ability to understand speech in quiet and noise using a listen-and-repeat task, but this test does not contain any sentences with different intonations or prosody changes. Since suprasegmental elements could also play a role in delivering messages, it is important to examine people’s ability to recognize the prosodic elements in speech. This study explores how prosodic elements are perceived by those with normal hearing (NH) and CIs.

Subjects and Methods

Participants

The inclusion criteria of the study were as follows: 1) native Korean speakers, 2) prelingual hearing loss, 3) speech intelligibility rating score of 5, and 4) Categories of Auditory Performance (CAP) scores of at least 6. The speech intelligibility rating assesses the intelligibility of speech produced by CI users [26]. It has five categories with category 1 being no intelligible speech and category 5 being intelligible speech with little or no concentration for the listener. The CAP also is a scale rating CI outcomes with seven categories with 0 being no display of sound awareness and 7 being able to perform a telephone conversation [27]. The exclusion criteria included individuals with neurological and mental disorders. All experimental procedures met the standards of the Helsinki Declaration of 2013 and were approved by Samsung Medical Center’s Institutional Review Board (IRB no. 2016-07-19). An informed consent document was obtained from the participants as well as their caregivers.

Speech perception testing

The Seoul National University Hospital (SNUH)-Speech Perception test which consists of 70 sentences (10 sentences per list × 7 lists) was used for testing [28]. The participants were asked to listen to the target stimuli, which was presented at 65 dB SPL, and repeat them back to the tester. The total score was calculated based on the correct responses (1 point for correct response and 0 point for incorrect response).

Speech prosody perception testing

The speech prosody perception test involved two types of testing. For all testing, recordings of stimuli were presented through a loudspeaker located 1 m away from the participants at 65 dB SPL. The first testing used three phonemes: /a/, /^/, and /i/. These phonemes have different meanings based on the prosody changes. For instance, depending on prosody change, /a/ can mean hurt (as in “ouch”) or opening of the mouth for eating. The participants were provided with pictures describing each phoneme stimulus and chose the right one for the stimulus they heard. The second part of the testing used words: positive word (PW) and neutral word (NW). As shown in Table 1, positive words have positive meaning and neutral words have neutral meaning. For all words, prosody changes (positive and negative) were made. For example, for the word “good job” that corresponds to PW, a negative prosody change was made so the meaning of the word changed from positive to negative. In the semantic-syntactic level, two combinations of words were made. For word combination (WC) 1, the adverb “really” was added to the verb or adjective belonging to each word group. For WC 2, the word itself was repeated twice. The participants listen to a total of 48 stimuli, consisting of PW and NW each. If they correctly identify whether the sound has positive or negative meanings, it is considered as a correct response. The percentage score is calculated by dividing the number of correct responses by the total number of stimuli.

Stimuli used for speech prosody perception testing

Pitch difference discrimination and melodic contour perception testing

The complex-tone pitch discrimination test from the Korean version of the Clinical Assessment of Music Perception (K-CAMP) was used for pitch difference discrimination testing [29]. The K-CAMP was administered using a MATLAB program (MathWorks, Natick, MA, USA) and sounds were presented through a loudspeaker located 1 m away from the participants. The complex-tone pitch discrimination is a two-alternative forced-choice test where the participants listened to two different tones and chose a tone that had higher pitch. Upon completion, the just-noticeable-difference limen (DL) in semitones were determined for 262, 330, and 392 Hz which are base frequencies.

The melodic contour perception testing was composed of melody contour discrimination (MCD) testing and melody contour identification (MCI) testing (Fig. 1). These tests were conducted based on Galvin, et al. [30] which introduced this test for CI user and NH listeners. For the MCD test, an ascending-descending melodic contour (A4-C#5-E5-C#5-A4) using the notes 1 (A4), 3 (C#5), and 5 (E5) within the range of A4 (440 Hz) to E5 (659 Hz) was created as a reference. Some of the constituent notes of the reference melodic contour were randomly adjusted by a semitone to create a pair of melodic contours: the reference contour and the adjusted contour (e.g., A4-C#5-E5-C#5-A4 vs. A4-C#5-E5-C#5-A4#). The first task was to discriminate between the reference contour and the adjusted contour pairs (MCD 1). If the participants perceived the two different, the second task was to identify the different one (MCD 2). MCI was also constructed based on A4 (440 Hz) as a reference. In MCI 1, participants listened to melody contour sounds and responded to which pair of contrasting contours (i.e., “rising” and “falling”) they corresponded to. In MCI 2, a picture panel of the nine melody contours was randomly presented, and participants were asked to identify the melody contour.

Fig. 1.

Melodic contour perception testing. A: Melody contour discrimination (MCD) testing. B: Melody contour identification (MCI) 1 and 2 testing.

Speech prosody production testing

For the speech prosody production testing, the task involved varying the emphasized syllable in a meaningless four-syllable sequence. The aim was to determine whether the participant produced appropriate prosodic variations or differences between the target syllable and the other syllables. In order to evaluate the participants’ ability to produce these syllables, the participants wore a headset microphone that was connected to a recording device. The microphone was positioned at a distance of 3–5 cm directly from the participants’ mouths. The audio signals were digitally recorded and saved as 16-bit WAV files with a sampling rate of 44.1 kHz. The consonant /m/ was used for this test. It is a nasal sound in Korean that does not involve articulatory involvement of the tongue, thus minimizing articulatory burden. Vowels were composed to reflect the presence or absence of distinctive features. The vowels /a/, /i/, and /u/, corresponding to the extreme points of the vowel chart in Korean, are used as reference points. Hence, the four-syllable meaningless sequences were composed by combining the consonant /m/ with a vowel: mamamama, mumumumu, and mimimimi.

Statistical analysis

Statistical analysis was performed with the Statistical Package for the Social Sciences version 26 (IBM Corp., Armonk, NY, USA). The Kruskal-Wallis test was performed to compare the groups’ performance. A p-value less than 0.05 was considered to be significant.

Results

Participant characteristics

A total of 26 participants (13 CI users and 13 age- and gender-matched controls) were recruited for the study (Table 2). The average ages of the participants were 19.8 years (SD=3.0) for the NH group and 19.0 years (SD=2.8) for the CI group. The CI users had prelingual hearing loss prior to implantation. The average age of implantation and CI use were 8.7 years (SD=3.04) and 10.3 years (SD=2.17), respectively. The average SIR and CAP scores were 5.0 (SD=0.1) and 7.0 (SD=0.3), respectively.

Participant characteristics

Speech prosody perception performance between the NH and CI groups

Table 3 illustrates the two groups’ performance in various prosody conditions. For both phoneme and words, the CI group showed significantly lower performance than the NH group in all conditions except in the NW negative prosody conditions for word and word combination 2.

Speech prosody perception performance between the groups

Pitch difference discrimination and melodic contour perception performance between the NH and CI groups

As described in Table 4, there was a statistically significant difference between the NH and CI groups on the complex-tone pitch discrimination test (p<0.05). Statistical significance was also observed for all tests (MCD 1 and 2 and MCI 1 and 2) on the melodic contour perception test (p<0.05 for MCI 1 and p<0.001 for the rest).

Complex-tone pitch discrimination in K-CAMP and melodic contour performance between the NH and CI groups

Speech perception performance between the NH and CI groups

All participants in the NH group showed 100% on test. The average score for the CI group was 89.8%—all participants scored higher than 75%. Statistical significance was observed for all test conditions (p<0.001) for the two groups.

Speech prosody production performance between the NH and CI groups

An analysis was conducted to examine whether there were differences in mean pitch, intensity, and duration (length) for each target syllable emphasized in the utterances “mamamama,” “mimimimi,” and “mumumumu” between the NH and CI groups. As indicated in Table 5, there were no distinct differences between the groups. There were also no significant differences observed between the target syllables with emphasis in “mamamama,” “mimimimi,” and “mumumumu.”

Speech prosody production performance between the NH and CI groups

Discussion

In this study, individuals’ ability to perceive prosodic elements was investigated. For the most part, as one could expect, NH listeners’ performance was significantly better than CI users. This not only includes the speech perception test, but also the speech prosody perception, pitch difference discrimination, and melodic contour perception tests. Although the CI users showed an average performance of 89.8% on the speech perception test, the average score for the NH group was significantly higher, which was 100%. At both phoneme and word levels, except for NW with negative prosody changes and WC 2 where the word was repeated twice, the NH group showed significantly better performance. The NH listeners were able to differentiate pitch and melodic contour significantly better than CI users. These findings are in line with previous studies that CI users are able to understand the segmental aspect of speech better after implantation, but still have difficulty with the prosodic elements embedded in speech [23,31-36]. Luo, et al. [35] explored emotion recognition ability in NH listeners and CI users using five emotions: angry, anxious, happy, sad, and neutral. The participants listened to recordings of speech and were asked to choose the emotion contained in the recording. While those with NH showed an average performance of 89.8%, the CI users showed significantly poor performance (44.9%), demonstrating that CI users have greater difficulty utilizing the suprasegmental aspect of speech. Nakata, et al. [23] recruited children with NH and CIs and compared their ability to perceive speech prosody. Sound stimuli containing neutral utterances with various prosody changes (happy, sad, and angry) were presented through a speaker and the children chose the emotion contained in the stimuli. The results showed that those with NH scored 100% while those using CIs performed poorly on the task (mean accuracy levels of 61% for happy, 64% for sad, and 39% for angry). For CI users with prelingual hearing loss, both segmental and suprasegmental information perception began through the use of CIs and other hearing devices as well as the aural rehabilitation process that focuses on the segmental aspect of speech. Therefore, during the early language development period in infancy, the CI group may have had relatively insufficient exposure to suprasegmental stimuli, such as the discrimination of suprasegmental information [37,38]. As a result, in the process of spoken communication, suprasegmental features might become confusing information that hinders the understanding of the intended meaning conveyed by the speaker, rather than serving as cues for speech perception. Even for those with postlingual hearing loss, they are able to maintain 85% of speech comprehension and a CAP score of 6 after receiving a CI, enabling them to do daily conversations without the need to visually observe lip movements. However, when subjected to the speech prosody testing proposed in this study, these individuals demonstrated poor performance in the range of 70% [39]. Consequently, while segmental information perception improved to a level comparable to individuals with NH through CIs, perception of subtle rhythmic variations remained limited. Furthermore, the participants’ performance in speech prosody testing was found to have a statistically significant inverse correlation with the duration of the time during which they did not receive any benefit from hearing aid use. Hence, even in individuals had sufficient opportunities to perceive speech sounds before language development, an extended period without the potential for auditory restoration though hearing aids has a negative impact on speech prosody perception. Moreover, it is predicted that even with CIs, when alterations in speech prosody contribute to changes in the original meaning, comprehension of spoken language remains challenging for CI users. For Korean, especially for spoken language, PWs are often used with a positive meaning, and words belonging to NW (i.e., so, where) are frequently used in everyday life in a negative sense when they are not used as interrogative pronouns [39,40]. In our study, as shown in Table 3, in the CI group, the median values for positive prosody condition were higher than the median values for negative prosody conditions in the PW task. On the other hand, in the NW task, the median values for the positive rhythmic conditions in both word task and word combination task were lower than the median values for the negative prosody conditions. Consequently, it can be expected that PW undergoes a transformation from its inherent positive meaning to a negative meaning, while the opposite occurs in NW, causing confusion in the perception of speech sound meaning and potentially having a negative impact. In terms of speech prosody production, no significant differences between the two groups were observed. These findings imply that even if the CI users can speak the sounds they perceive, it should not be assumed that they have understood them correctly.

In sum, findings of this study emphasize the need for an assessment tool focusing on prosodic elements of speech in clinical environments—even if audibility can be restored through CIs, an important element, prosody, needs to be assessed as well in order to accurately evaluate one’s speech performance. In addition, development of a CI signal processing algorithm that takes suprasegmental factors of speech sound into consideration is needed. As prosody is a critical element to deliver and perceive emotions, CI users’ ability to perceive and produce the prosodic element needs to be thoroughly assessed and this will also help with their aural rehabilitation plans. Subsequent studies with larger sample sizes and a larger variety of participant characteristics are necessary to further examine the need and potential of the inclusion of the prosodic element in audiological assessment and rehabilitation.

Acknowledgements

We thank all individuals participated in this study.

Notes

Conflicts of Interest

The authors have no financial conflicts of interest.

Author Contributions

Conceptualization: Eun Yeon Kim. Data curation: Hye Yoon Seol. Formal analysis: Eun Yeon Kim, Hye Yoon Seol. Investigation: Eun Yeon Kim, Hye Yoon Seol. Methodology: Eun Yeon Kim. Project administration: Eun Yeon Kim, Hye Yoon Seol. Resources: Eun Yeon Kim, Hye Yoon Seol. Software: Eun Yeon Kim, Hye Yoon Seol. Supervision: Hye Yoon Seol. Validation: Eun Yeon Kim, Hye Yoon Seol. Visualization: Eun Yeon Kim, Hye Yoon Seol. Writing—original draft: Eun Yeon Kim. Writing—review & editing: Eun Yeon Kim, Hye Yoon Seol. Approval of final manuscript: Eun Yeon Kim, Hye Yoon Seol.

References

1. Jakobson R. Linguistics and poetics. In : Sebeok TA, ed. Style in Language Cambridge, MA: MIT Press; 1960. p. 350–77.
2. Norris D. Shortlist: a connectionist model of continuous speech recognition. Cognition 1994;52:189–234.
3. Keightley ML, Winocur G, Graham SJ, Mayberg HS, Hevenor SJ, Grady CL. An fMRI study investigating cognitive modulation of brain regions associated with emotional processing of visual stimuli. Neuropsychologia 2003;41:585–96.
4. Critchley H, Daly E, Phillips M, Brammer M, Bullmore E, Williams S, et al. Explicit and implicit neural mechanisms for processing of social information from facial expressions: a functional magnetic resonance imaging study. Hum Brain Mapp 2000;9:93–105.
5. Haugen E. Phoneme or prosodeme? Language 1949;25:278–82.
6. Varadarajan VV, Sydlowski SA, Li MM, Anne S, Adunka OF. Evolving criteria for adult and pediatric cochlear implantation. Ear Nose Throat J 2021;100:31–7.
7. Seol HY, Moon IJ. Hearables as a gateway to hearing health care. Clin Exp Otorhinolaryngol 2022;15:127–34.
8. Chang SO, Koo JW. Cochlear implant in children. Korean J Audiol 1997;1:13–21.
9. Cho YB. Basic consideration for cochlear implant. Korean J Audiol 1997;1:3–12.
10. Hainarosie M, Zainea V, Hainarosie R. The evolution of cochlear implant technology and its clinical relevance. J Med Life 2014;7 Spec No. 2:1–4.
11. Ketterer MC, Häussler SM, Hildenbrand T, Speck I, Peus D, Rosner B, et al. Binaural hearing rehabilitation improves speech perception, quality of life, tinnitus distress, and psychological comorbidities. Otol Neurotol 2020;41:e563–74.
12. Laske RD, Veraguth D, Dillier N, Binkert A, Holzmann D, Huber AM. Subjective and objective results after bilateral cochlear implantation in adults. Otol Neurotol 2009;30:313–8.
13. Moon IJ, Kim EY, Jeong JO, Chung WH, Cho YS, Hong SH. The influence of various factors on the performance of repetition tests in adults with cochlear implants. Eur Arch Otorhinolaryngol 2012;269:739–45.
14. Chang YS, Hong SH, Kim EY, Choi JE, Chung WH, Cho YS, et al. Benefit and predictive factors for speech perception outcomes in pediatric bilateral cochlear implant recipients. Braz J Otorhinolaryngol 2019;85:571–7.
15. Lee JY, Hong SH, Moon IJ, Kim EY, Baek E, Seol HY, et al. Effect of cochlear implant electrode array design on electrophysiological and psychophysical measures: lateral wall versus perimodiolar types. J Audiol Otol 2019;23:145–52.
16. Tyagi P, Chauhan D, Singh A, Bhutada M, Sikka K, Chaudhary T, et al. Clinical and social outcomes of cochlear implantation in older prelinguals. J Audiol Otol 2023;27:63–70.
17. Šantić IŠ, Bonetti L. Language intervention instead of speech intervention for children with cochlear implants. J Audiol Otol 2023;27:55–62.
18. Hwang K, Lee JY, Oh HS, Lee BD, Jung J, Choi JY. Feasibility of revision cochlear implant surgery for better speech comprehension. J Audiol Otol 2019;23:112–7.
19. Marx M, James C, Foxton J, Capber A, Fraysse B, Barone P, et al. Speech prosody perception in cochlear implant users with and without residual hearing. Ear Hear 2015;36:239–48.
20. Green T, Faulkner A, Rosen S. Enhancing temporal cues to voice pitch in continuous interleaved sampling cochlear implants. J Acoust Soc Am 2004;116(4 Pt 1):2298–310.
21. Peng SC, Tomblin JB, Turner CW. Production and perception of speech intonation in pediatric cochlear implant recipients and individuals with normal hearing. Ear Hear 2008;29:336–51.
22. Peng SC, Tomblin JB, Cheung H, Lin YS, Wang LS. Perception and production of mandarin tones in prelingually deaf children with cochlear implants. Ear Hear 2004;25:251–64.
23. Nakata T, Trehub SE, Kanda Y. Effect of cochlear implants on children’s perception and production of speech prosody. J Acoust Soc Am 2012;131:1307–14.
24. Klieve S, Jeanes RC. Perception of prosodic features by children with cochlear implants: is it sufficient for understanding meaning differences in language? Deaf Educ Int 2001;3:15–37.
25. Lenden JM, Flipsen P Jr. Prosody and voice characteristics of children with cochlear implants. J Commun Disord 2007;40:66–81.
26. Wilkinson AS, Brinton JC. Speech intelligibility rating of cochlear implanted children: inter-rater reliability. Cochlear Implants Int 2003;4:22–30.
27. Archbold S, Lutman ME, Nikolopoulos T. Categories of auditory performance: inter-user reliability. Br J Audiol 1998;32:7–12.
28. The Korean Audiological Society. Speech perception test in practice 2nd edth ed. Seoul: Hakjisa; 2021. p. 281–5.
29. Jung KH, Cho YS, Cho JK, Park GY, Kim EY, Hong SH, et al. Clinical assessment of music perception in Korean cochlear implant listeners. Acta Otolaryngol 2010;130:716–23.
30. Galvin JJ 3rd, Fu QJ, Shannon RV. Melodic contour identification and music perception by cochlear implant users. Ann N Y Acad Sci 2009;1169:518–33.
31. Govaerts PJ, De Beukelaer C, Daemers K, De Ceulaer G, Yperman M, Somers T, et al. Outcome of cochlear implantation at different ages from 0 to 6 years. Otol Neurotol 2002;23:885–90.
32. Tobey EA, Hasenstab S. Effects of a nucleus multichannel cochlear implant upon speech production in children. Ear Hear 1991;12(4 Suppl):48S–54S.
33. Most T, Peled M. Perception of suprasegmental features of speech by children with cochlear implants and children with hearing AIDS. J Deaf Stud Deaf Educ 2007;12:350–61.
34. Pereira C. The perception of vocal affect by cochlear implantees. In : Waltzman SB, Cohen N, eds. Cochlear Implants New York: Thieme; 2000. p. 343–5.
35. Luo X, Fu QJ, Galvin JJ 3rd. Vocal emotion recognition by normalhearing listeners and cochlear implant users. Trends Amplif 2007;11:301–15.
36. Rastegarianzadeh N, Shahbodaghi M, Faghihzadeh S. Study of phonological awareness of preschool and school aged children with cochlear implant and normal hearing. Korean J Audiol 2014;18:50–3.
37. Grossmann T, Striano T, Friederici AD. Infants’ electric brain responses to emotional prosody. Neuroreport 2005;16:1825–8.
38. Ahn SJ, Kim YJ, Choung YH, Park KH, Moon SK. The perception and production of the final question intonations for the cochlear implantee. Korean J Audiol 2003;7:136–40.
39. Kim EY, Moon IJ, Cho YS, Chung WH, Hong SH. The relationship between perception of prosody, pitch discrimination, and melodic contour identification in cochlear implants recipients. J Music Hum Behav 2017;14:1–18.
40. Lee HG. The pragmatics of the discourse particle eti in Korean. Urimalgeul: Korean Lang Lit 2008;44:83–111.

Article information Continued

Fig. 1.

Melodic contour perception testing. A: Melody contour discrimination (MCD) testing. B: Melody contour identification (MCI) 1 and 2 testing.

Table 1.

Stimuli used for speech prosody perception testing

Meaning Conditions
Word Word combination 1 Word combination 2
PW Good job Really good Good job, good job
Joyful Really joyful Joyful, joyful
Fun Really fun Fun, fun
Cool Really cool Cool, cool
Excited Really excited Excited, excited
Happy Really happy Happy, happy
Pretty Really pretty Pretty, pretty
Good Really good Good, good
NW Thus Really thus Thus, thus
Who Really who Who, who
When Really when When, when
Where Really where Where, where

PW, positive word; NW, neutral word

Table 2.

Participant characteristics

No. NH (n=13)
No. CI (n=13)
Sex Age (yr) Sex Age (yr) CI age CI use CI side Device
NH1 F 15.1 CI1 F 16.2 7.0 9.2 L Harmony
NH2 M 20.2 CI2 M 20.4 11.4 8.9 L Harmony
NH3 F 23.9 CI3 F 22.1 12.1 10.0 R Freedom
NH4 F 19.0 CI4 F 17.7 11.4 6.3 L Freedom
NH5 M 16.0 CI5 M 16.9 2.8 14.1 L Freedom
NH6 F 21.1 CI6 F 19.2 8.1 11.1 R Freedom
NH7 F 17.7 CI7 F 17.2 3.6 13.6 R Freedom
NH8 M 19.1 CI8 M 16.9 8.9 8.0 R Freedom
NH9 M 24.2 CI9 M 23.1 10.9 12.1 R Auria
NH10 M 24.9 CI10 M 24.7 13.2 11.5 R Auria
NH11 M 19.0 CI11 M 16.8 7.6 9.2 R Freedom
NH12 M 18.2 CI12 M 17.4 9.2 8.2 R Freedom
NH13 M 19.0 CI13 M 17.9 6.7 11.2 L Auria

NH, normal hearing; CI, cochlear implant

Table 3.

Speech prosody perception performance between the groups

Prosody condition Presentation (%)
Phoneme
Word
WC 1
WC 2
NH CI NH CI NH CI NH CI
Positive & negative PW 100* (100-100) 100 (20-100) NA NA NA NA NA NA
 Positive NA NA 100* (100-100) 100 (50-100) 100** (100-100) 100 (75-100) 100** (75-100) 75 (25-100)
 Negative NA NA 100** (100-100) 50 (0-100) 100** (100-100) 75 (25-100) 100** (75-100) 75 (50-100)
NW
 Positive NA NA 100** (75-100) 100 (0-100) 100** (75-100) 75 (25-100) 100** (75-100) 50 (25-100)
 Negative NA NA 100 (75-100) 100 (50-100) 100** (100-100) 75 (25-100) 100 (100-100) 100 (75-100)

Values are presented as median (min-max).

*

p<0.05;

**

p<0.001.

PW, positive word; NW, neutral word; NH, normal hearing; CI, cochlear implant; NA, not available; WC, word combination

Table 4.

Complex-tone pitch discrimination in K-CAMP and melodic contour performance between the NH and CI groups

Factor of prosody Group
NH CI
Complex-tone pitch discrimination (DL)
 Threshold
  260 Hz 1.22 (0.50-3.44)* 2.39 (0.56-8.11)
  330 Hz 0.78 (0.50-2.50)* 1.72 (0.50-11.50)
  391 Hz 0.83 (0.50-1.72)* 3.60 (0.50-11.00)
Melodic contour perception (%)
 MCD 1 100 (100-100)** 85.71 (57.14-100)
 MCD 2 100 (85.71-100)** 42.86 (14.29-71.43)
 MCI 1 100 (100-100)* 100 (0-100)
 MCI 2 100 (94.44-100)** 61.11 (11.11-100)

Values are presented as median (min-max).

*

p< 0.05;

**

p<0.001.

NH, normal hearing; CI, cochlear implant; DL, just-noticeable-difference limen; MCD, melodic contour discrimination; MCI, melodic contour identification

Table 5.

Speech prosody production performance between the NH and CI groups

NH
CI
Pitch (Hz) Intensity (dB) Length (sec) Pitch (Hz) Intensity (dB) Length (sec)
ma1mamama 146.92 (128.63-281.03) 63.22 (48.65-75.27) 0.44 (0.35-0.64) 178.15 (125.17-321.05) 68.56 (52.11-79.03) 0.44 (0.33-0.5)
mama2mama 161.82 (128.81-283.49) 64.98 (52.27-76.08) 0.43 (0.31-0.51) 203.97 (125.95-357.50) 69.38 (57.46-76.15) 0.40 (0.27-0.49)
mamama3ma 158.84 (125.38-289.40) 66.66 (53.00-75.35) 0.44 (0.31-0.54) 207.10 (125.55-297.32) 68.03 (56.19-74.59) 0.38 (0.29-0.50)
mamamama4 150.94 (114.86-272.07) 64.02 (51.9-71.28) 0.37 (0.22-0.54) 189.99 (125.85-313.26) 67.69 (53.96-74.64) 0.33 (0.22-0.51)
mi1mimimi 164.32 (132.26-292.65) 66.96 (52.69-73.00) 0.44 (0.25-0.55) 201.90 (162.34-333.02) 69.18 (49.27-76.86) 0.42 (0.30-0.54)
mimi2mimi 160.52 (133.14-305.77) 73.24 (52.38-78.32) 0.45 (0.28-0.52) 242.78 (161.57-359.82) 69.41 (49.30-77.68) 0.39 (0.27-0.51)
mimimi3mi 157.76 (127.70-288.79) 58.34 (46.50-71.76) 0.44 (0.27-0.51) 186.28 (136.61-321.17) 69.40 (51.23-76.37) 0.42 (0.29-0.58)
mimimimi4 140.61 (114.59-282.56) 8.93 (2.14-25.76) 0.43* (0.32-0.6) 217.97 (154.57-294.28) 67.56 (54.11-71.63) 0.34 (0.27-0.49)
mu1mumumu 146.04 (133.65-294.56) 69.08 (51.96-76.09) 0.45 (0.33-0.49) 227.42 (152.19-333.21) 71.38 (51.97-76.23) 0.41 (0.25-0.55)
mumu2mumu 149.54 (133.92-292.38) 70.39 (51.62-73.53) 0.46 (0.31-0.49) 241.90 (144.42-375.37) 70.42 (53.67-74.97) 0.41 (0.25-0.58)
mumumu3mu 162.41 (130.19-296.35) 64.50 (51.77-75.6) 0.41 (0.27-0.59) 240.51 (154.08-359.18) 70.78 (53.43-77.53) 0.46 (0.27-0.54)
mumumumu4 146.29 (108.49-290.77) 68.16 (51.04-74.1) 0.41 (0.23-0.69) 215.95 (137.73-315.29) 69.53 (53.21-73.35) 0.38 (0.26-0.45)

Values are presented as median (min-max). NH, normal hearing; CI, cochlear implant