PERCEPTION AND CONFUSION OF SPEECH SOUNDS BY CHILDREN AND ADULTS WITH A COCHLEAR IMPLANT Arne K. Rødvik Rikshospitalet University Hospital Norway Aim A Norwegian non-word test, will be used to identify typical confusions of consonants and vowels by children and adults with a cochlear implant. We expect that this study will be beneficial both for the fitting process of the cochlear implants, and for the work of the speech therapists post operatively. Moreover, it may be useful for the improvement of the speech processing algorithms of the cochlear implants. 1 Description of the speech perception test • Developed at the University of Iowa in USA and adapted to Norwegian at Rikshospitalet • Contains two consonant confusion tests and one vowel confusion test • The words in the test are being read by a 25 years old woman Description of the speech perception test (cont.) /! C! / /! C! / /bVb/ /! ! ! ! ! /! ! ! ! ! /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / /! ! ! ! / 2 Method The informants will be asked to listen to the recorded nonwords aCa, iCi and bVb, and asked to repeat what they hear. Their repetitions will be recorded and transcribed phonetically. The collected data will be arranged in confusion matrices to find the most common and the most uncommon speech sound confusions. The matrices will be analyzed by counting the correct repetitions, as well as by phonetic analysis. An acoustic analysis of the recordings will also be performed, in order to look at characteristic features of the voice quality of the cochlear implanted informants. Test equipment and setup • Pioneer DVD player and Sony APM-X5A loudspeakers. • The distance between the informant and the middle point between the loudspeakers is ca 2 meters. • The average loudness in listening position is 65 dB (SPL). • The words stored on the DVD are represented by bar codes, and are played one by one by means of a bar code reader connected to the DVD player. • The utterances of the informants are recorded to a Sony NZ-707 MiniDisc-player and a Sony ECM-MS907 stereo microphone. • The distance between the microphone and the mouth of the informant is 50 cm. • The testing is performed in an anechoic chamber. 3 Inclusion criteria of the informants • Post lingually profoundly deaf adults with a cochlear implant (CI) • Speech understanding on IOWAsentences10 above 50%. • Speech understanding on a test with monosyllabic words above 50 % • Ability to repeat nonsens words. • Ability to pronounce all of the Norwegian phonemes11 Briefing of the informants Before we run the test, the informant are told that: • The words are nonsens words • Whether s/he is to listen for consonants or vowels • Which sound comes before and after the significant sound • S/he should make a guess if unsecure of which sound is presented • A response must be given to all of the presented words 4 Confusion matrices Pilot study A pilot study with 5 informants has been conducted9, and the results showed that • Manner of articulation is rarely confused • Place of articulation is often confused Manner of articulation Plosives Frikatives Nasals Liquids Place of articulation Bilabial Labio dental /p/, /b/ /f/, /V/ /m/ Dental/alveolar/ Palatal postalveolar /t/, /d/ /s/, /S/ /ç/, /Æ/ /n/ Velar /k/, /g/ /N/(ng-lyd) /r/, /l/ 5 Pilot study (cont.) • There was a tendency that consonants were more often confused in the /iCi/ combination than in the /ACA/ combination. • Voiced and unvoiced sounds were rarely confused • Nasal and not nasal sounds were rarely confused, except /l/ and /n/ • Words which resemble meaningful words were more easily recognised than nonsens words Changes of the protocol from the pilot study • More detailed phonetic transcription • Randomization of the informants’ responses before transcription • Larger number of informants • Inclusion of children 6 Hypotheses Based on the results of the pilot study, the following hypotheses will be investigated further: • The confusion of speech sounds with the same place of articulation and different manner of articulation is more common than the confusion of speech sounds with the same manner and different place of articulation. • Consonants are more often confused in the iCi context than in the aCa context. • Nasal and oral speech sounds are seldomly confused, except /l/ and /n/. • Nonwords which resemble real words are more easily recognized than nonwords which do not resemble real words. • Postlingually and prelingually deaf confuse speech sounds differently. • The voice quality of the cochlear implanted is proportional to the ability to discriminate between speech sounds. • Factors that make a significant influence on the ability to recognize nonwords are time period of using the CI, time of deafness prior to the implantation, amount and quality of speech and listening training post operatively, type of implant and speech processing strategy of the implant. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Bench, J., Kowal, A. and Bamford, J. M. (1979). The BKB (Bamford-Kowal-Bench) sentence lists for partially hearing children. Br J Audiol, 13, 108-112. Blamey, P. J. and Clark, G. M. (1990). Place coding of vowel formants for cochlear implant patients. J. Acoust. Soc. Am, 88 (2). 667-673. Dubno, J. R. and Levitt, H. (1980). Predicting consonant confusions from acoustic analysis. J. Acoust. Soc. Am, 69 (1). 249-261. Kent, R. D. and Read, C. (1992). The Acoustic Analysis of Speech. San Diego: Singular Publishing. Kristoffersen, G. (2000). The phonology of Norwegian. Oxford University press. Miller, G. A. and Nicely, P. E. (1955). An Analysis of Perceptual Confusions Among Some English Consonants. J. Acoust. Soc. Am, 27 (2), 338-352. Munson, B., Donaldson, G. S., Allen, S. L., Collison, E. A. and Nelson, D. A. (2003). Patterns of phoneme perception errors by listeners with cochlear implants as a function of overall speech perception ability. J. Acoust. Soc. Am, 113 (2), 925-935. Mülder, H. E., van Olphen, A. F., Bosman, A. and Smoorenburg, G. F. (1992). Phoneme Recognition by Deaf Individuals Using the Multichannel Nucleus Cochlear Implant. Acta otolaryngol (Stockh), 112, 946-955. Rødvik, A. K. (2008). Perception and confusion of speech sounds by adults with a cochlear implant. Clinical Linguistics & Phonetics 22(4-5), 371-378. Teig, E., Lindeman, H. H., Tvete, O., Hanche-Olsen, S. and Rasmussen, K. (1993). Audiovisual Test Programs in Native Languages. Advances in Oto-Rhino-Laryngology, 48, 199-202. Tingleff, H. and Tingleff, Ø. (2002). Norsk fonemtest. Damm forlag. Tyler, R. S., Preece, J. P. and Tye Murray, N (1987). Iowa Audiovisual Speech Perception Laser Videodisc. Laser Videodisc and Laboratory Report, University of Iowa, Department of Otolaryngology Head and Neck Surgery. van Wieringen, A. and Wouters, J. (1999). Natural Vowel and Consonant Recognition by Laura Cochlear Implantees. Ear and Hearing, 20 (2), 89-103. Wie, O. B., Falkenberg, E. S., Tvete, O. and Tomblin, B. (2007). Children with a cochlear implant: Characteristics and determinants of speech recognition, speech-recognition growth rate, and speech production. International Journal of Audiology, 46, 232-243. 7 Thanks for your attention! 8
© Copyright 2026 Paperzz