Rehabilitation of speech disorders following glossectomy, based on ultrasound visual illustration and feedback

ABSTRACT

Intraoral surgery for tongue cancer usually induces speech disorders that have a negative impact on communication and quality of life. Studies have documented the benefit of tongue ultrasound imaging as a visual articulatory feedback for speech rehabilitation. This study aims to assess specifically the complementary contribution of visual feedback to visual illustration (i.e. the display of ultrasound video of target language movements) for the speech rehabilitation of glossectomised patients. Two therapy conditions were used alternately for ten glossectomised French patients randomly divided into two cohorts. The IF cohort benefitted from 10 sessions using illustration alone (IL condition) followed by 10 sessions using illustration supplemented by visual feedback (IL+F condition). The FI cohort followed the opposite protocol, i.e. the first 10 sessions with the IL+F condition, followed by 10 sessions with the IL condition. Phonetic accuracy (Percent Consonants Correct) was monitored at baseline (T0, before the first series) and after each series (T1 and T2) using clinical speech-language assessments. None of the contrasts computed between the two conditions, using logistic regression with random effects models, were found to be statistically significant for the group analysis of assessment scores. Results were significant for a few individuals, with balanced advantages in both conditions. In conclusion, the use of articulatory visual feedback does not seem to bring a decisive advantage over the use of visual illustration, though speech therapists and patients reported that ultrasound feedback was useful at the beginning. This result should be confirmed by similar studies involving other types of speech disorders.

KEYWORDS:

• Speech therapy
• tongue ultrasound imaging
• visual illustration
• visual feedback
• articulation disorder
• assistive technologies

Acknowledgments

This work was funded by the Rhône-Alpes Region ARC6 program ‘Information and Communication Technologies and Innovative IT Practices’, as part of Diandra Fabre’s doctoral thesis. Authors are very indebted to SLP Bérengère Gal who shared the rehabilitation work, to SLP Ondine Champavère who performed the second assessment based on audio recordings, and to Dr Thomas Dell’Accio who participated in the follow-up examinations.

The authors are very grateful to the editor and the two reviewers for their relevant comments, which contributed significantly to the improvement of the article.

Declaration of interest

The authors report no conflicts of interest.

Notes

¹ This software can be freely downloaded at http://www.ultraspeech.com.

² This database can be freely downloaded at http://www.ultraspeech.com.

³ http://www.articulateinstruments.com.

⁴ This software can be freely downloaded at http://www.ultraspeech.com.

Additional information

Funding

This work was supported by the Rhône-Alpes Region ARC6 program “information and communication technologies and innovative it practices” [ARC6 - 2013 - 13-009712-01].