Abstract
Objective: This study investigated whether adding an additional modality, namely ultrasound tongue imaging, to perception-based phonetic transcription impacted on the identification of compensatory articulations and inter-rater reliability.
Patients and Methods: Thirty-six children aged 3 to 12 with CLP were recorded producing repetitions 5 of /aCa/ for all places of articulation with simultaneous audio and probe-stabilised ultrasound. Three types of transcriptions were performed: 1. Traditional phonetic transcription by a CLP specialist from 7 audio recording; 2. Qualitative observations from the live ultrasound by the clinician recording the data; and, 3. Qualitative ultrasound-aided transcription by two ultrasound-trained clinicians. We compared the number of consonants identified as in error by each transcriber and then classified 10 errors into eight different subcategories.
Results: Both the ultrasound-aided and traditional transcriptions yielded similar error-detection rates, however these were significantly higher than the observations recorded live in the clinic. Inter-rater reliability for the ultrasound transcribers was substantial, compared to moderate for the traditional transcribers. Ultrasound-aided transcribers were more likely to identify covert errors such as double articulations and retroflexion than the audio-only transcribers.
Conclusion: Ultrasound-tongue imaging is a useful complement to traditional phonetic transcription.
Patients and Methods: Thirty-six children aged 3 to 12 with CLP were recorded producing repetitions 5 of /aCa/ for all places of articulation with simultaneous audio and probe-stabilised ultrasound. Three types of transcriptions were performed: 1. Traditional phonetic transcription by a CLP specialist from 7 audio recording; 2. Qualitative observations from the live ultrasound by the clinician recording the data; and, 3. Qualitative ultrasound-aided transcription by two ultrasound-trained clinicians. We compared the number of consonants identified as in error by each transcriber and then classified 10 errors into eight different subcategories.
Results: Both the ultrasound-aided and traditional transcriptions yielded similar error-detection rates, however these were significantly higher than the observations recorded live in the clinic. Inter-rater reliability for the ultrasound transcribers was substantial, compared to moderate for the traditional transcribers. Ultrasound-aided transcribers were more likely to identify covert errors such as double articulations and retroflexion than the audio-only transcribers.
Conclusion: Ultrasound-tongue imaging is a useful complement to traditional phonetic transcription.
| Original language | English |
|---|---|
| Number of pages | 1 |
| Publication status | Published - 3 Apr 2019 |
| Event | Craniofacial Society of Great Britain and Ireland Annual Scientific Meeting: Long Term Outcomes - London, United Kingdom Duration: 10 Apr 2019 → 12 Apr 2019 https://craniofacialconference19.co.uk/ |
Conference
| Conference | Craniofacial Society of Great Britain and Ireland Annual Scientific Meeting |
|---|---|
| Country | United Kingdom |
| City | London |
| Period | 10/04/19 → 12/04/19 |
| Internet address |
Fingerprint
Keywords
- cleft lip and palate
- phonetic transcription
- speech outcomes
Cite this
Cleland, J., Lloyd, S., Crampin, L., Campbell, L., Sugden, E., Zharkova, N., & Palo, P. (2019). Improving the reliability of phonetic transcription in cleft lip and palate using ultrasound tongue imaging. Poster session presented at Craniofacial Society of Great Britain and Ireland Annual Scientific Meeting, London, United Kingdom.