You are viewing the SNL 2018 Archive Website. For the latest information, see the Current Website.

Poster C44, Friday, August 17, 10:30 am – 12:15 pm, Room 2000AB

Assessing speech movements in people who stutter using real-time MRI of the vocal tract

Charlotte Wiltshire1, Jennifer Chesters1, Mark Chiew1, Kate E Watkins1;1University of Oxford

Introduction: The speech motor systems of people who stutter (PWS) are less stable than those of fluent controls. Previous measures of lip aperture over repeated utterances of nonwords reported greater variability in PWS than controls (Smith et al., 2010); variability increased with the length and phonological complexity of the nonword and decreased over repetitions in PWS. Here, we examined these effects in PWS using a novel technique of real-time MRI of the vocal tract during speech. Method: Mid-sagittal images of the vocal tract from lips to larynx were acquired with in-plane spatial resolution of 2mm x 2mm using a radial FLASH sequence (TE/TR = 1.4/2.5ms) with golden angle sampling. Images were reconstructed at 33.3 frames per second. We scanned seven adult men with moderate to severe stuttering (mean age = 27.9 years; range = 19-40) and five age-matched fluent male controls (mean age = 27.6 years; range = 20-41). Participants repeated nonwords at their normal speaking rate during scanning. Stimuli were identical to those used by Smith et al (2010); four nonwords of increasing length and phonological complexity: “mab” (/mæb/), “mabshibe” (/mæbʃaIb/), “mabfieshabe” (/mæbfaIʃeIb/), and “mabshaytiedoib” (/mæbʃeitaIdɔIb/) and a fifth nonword, “mabteebeebee” (/mæbtibibi/), matched for length but with low phonological complexity. Nonwords were presented in a random order. The imaging data were analysed using a custom Matlab toolbox that uses air-tissue boundary segmentation to track precise movements within the vocal tract (Kim et al., 2014). In this preliminary analysis, lip aperture (the distance between the upper and lower lips in mm) was measured across the entire utterance of the one- and four-syllable nonwords. The size of the lip movement for each nonword was calculated by summing the measurements across all frames to capture variability in both space and time. Variability across repeated utterances was calculated using the coefficient of variation. The change in coefficient of variation from the first five repetitions to the last five repetitions was also calculated. Group comparisons were made used a Mann-Whitney U test. Results: PWS produced larger movements than controls for both the complex (U(8) = 24, Z = 2.4, p = .016) and simple (U(9) = 27, Z = 2.2, p = .03) four-syllable nonwords but showed no difference in the overall size of the speech movement for the one-syllable nonword. There was no difference between PWS and controls in the coefficients of variation for one- and four-syllable nonwords. The coefficient of variation also did not change significantly from the first five to the last five utterances of these nonwords in either group and the size of this change did not differ between PWS and controls. Conclusion: Using real-time MRI of the vocal tract, we found that PWS produced larger speech movements than controls when speaking multi- but not mono-syllabic nonwords. This result is consistent with previous findings in PWS (Max, Caruso & Gracco, 2003). However, this preliminary analysis of a small sample revealed no differences between PWS and fluent controls in the variability of lip movements.

Topic Area: Language Disorders