Using ultrasound to treat vowel disorders: A single case study

Roxburgh, Zoe and Heyde, Cornelia and Capps, Matthew and Scobbie, James and Cleland, Joanne (2016) Using ultrasound to treat vowel disorders: A single case study. In: International Clinical Phonetics and Linguistics Association Conference, 2016-06-14 - 2016-06-18, Dalhousie University. (

Full text not available in this repository.Request a copy


Children acquiring English will demonstrate vowel errors mostly between age 1;6 and 2;0 (Hare, 1983; Paschall, 1983) and will have developed an almost complete vowel system (93.3%) by age three (Ball & Gibbon, 2013). Vowel acquisition can be problematic for some children leading to vowel error patterns, such as lowering and fronting. This can result in a perceptual collapse in vowel contrast; context sensitive processes where surrounding consonants influence the vowel; and more unusual idiosyncratic systems (Reynolds, 1990; Reynolds, 2013; Pollock, 2013; Bates, Watson, & Scobbie, 2013). Robb, Beleile, and Lee (1993) and Speake, Stackhouse, and Pascoe (2012) provide encouragement that both vowel-focused and unfocused speech and language therapy can increase vowel inventory and accuracy of production. Ultrasound Visual Bio-Feedback (U-VBF) Therapy has become increasingly popular for the remediation of persistent speech sound disorders (SSDs), however there are few studies investigating the use of U-VBF therapy for the remediation of vowel disorders. Current evidence for the remediation of vowels is within the hearing impairment population, showing limited evidence of improvement (Klajman et al., 1988; Wein et al., 1991; Bacsfalvi et al., 2007). The UltraPhonix Project aims to investigate the effectiveness of ultrasound therapy for 20 children aged six to 16 with persistent SSDs. This study uses a single-subject multiple baseline design with Michael (pseudonym),aged 6;5 years, presenting with a vowel disorder. Our research questions were: 1. Does a course of ultrasound visual biofeedback treatment improve accuracy of the targeted phoneme(s) in a. Words/pseudo-words and phrases trained during the therapy (treated word lists)? b. Words/pseudo-words and phrases not trained during the therapy (untreated wordlists)? 2. Are listeners able to perceive a difference in the production of /ɛ/pre- and post-therapy tokens? 3. Does acoustic and instrumental data provide evidence of covert errors/contrasts? Repeated measures of the Phonology subtest of the Diagnostic Evaluation of Articulation and Phonology (DEAP) (Dodd et al, 2002) and an untreated wordlist containing X tokens of /ɪ ɛ a ɑ/ were undertaken across three baseline assessments. Michael received 10 therapy sessions using ultrasound visual biofeedback (U-VBF) targeting /ɛ/. Assessments were repeated mid-therapy and will be repeated post therapy and at a maintenance session three months post-therapy to test for generalisation. Ultrasound datawas aquired using an Ultrasonix® SonixRP machine remotely controlled via Ethernet from a PC running Articulate Assistant Advanced™ (AAA) software (Articulate Instruments 2010). A probe stabilising headset is being used, to ensure accurate measurements are gathered. To ensure that headset movement is accounted for, a video from a headset-mounted micro-camera is being used, which also captures lip data. A headset-mounted microphone is also being used to record audio data. Therapy is currently on-going. We will present a fine phonetic analysis of Michael’s vowel system including perceptual and instrumental findings to investigate possible covert errors or contrasts. Outcome measures will be based on auditory judgements by trained and untrained listeners (phonetic transcription and perceptual evaluation or pre- and post-therapy measures), acoustic analysis (formant analysis) and articulatory data (tongue surface shape analysis).