Comparing metrics for quantification of children’s tongue shape complexity using ultrasound imaging

ABSTRACT

Speech sound disorders can pose a challenge to communication in children that may persist into adulthood. As some speech sounds are known to require differential control of anterior versus posterior regions of the tongue body, valid measurement of the degree of differentiation of a given tongue shape has the potential to shed light on development of motor skill in typical and disordered speakers. The current study sought to compare the success of multiple techniques in quantifying tongue shape complexity as an index of degree of lingual differentiation in child and adult speakers. Using a pre-existing data set of ultrasound images of tongue shapes from adult speakers producing a variety of phonemes, we compared the extent to which three metrics of tongue shape complexity differed across phonemes/phoneme classes that were expected to differ in articulatory complexity. We then repeated this process with ultrasound tongue shapes produced by a sample of young children. The results of these comparisons suggested that a modified curvature index and a metric representing the number of inflection points best reflected small changes in tongue shapes across individuals differing in vocal tract size. Ultimately, these metrics have the potential to reveal delays in motor skill in young children, which could inform assessment procedures and treatment decisions for children with speech delays and disorders.

KEYWORDS:

• Speech development
• lingual differentiation
• ultrasound imaging

Acknowledgments

We thank Katherine M. Dawson for sharing the data that formed the basis of the adult analyses, Graham Tomkins Feeny for coordination of measurement in GetContours, and Emily Phillips, Megan Leece, and Twylah Campbell for data collection at the three sites. We acknowledge Siemens Medical Solutions USA, Inc., for making their Acuson ultrasound scanner available.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.

Notes

¹ Motor maturation is not the only factor that influences the sequence of phoneme acquisition; factors such as auditory distinctness and frequency also undoubtedly play a role. Thus, we cannot expect tongue shape complexity measures to provide a comprehensive account of all acquisition phenomena. However, we still consider it a worthwhile enterprise in light of the important role of motor factors in shaping speech acquisition.

² This assumption may be questioned based on reports of vowel errors in typically developing children as old as 30 months (Otomo & Stoel-Gammon, 1992). However, data from Kent and Rountrey (2020) suggest that the cardinal vowels are in place by age two, including the low front vowel /æ/, which is used as a reference vowel in the present study.

³ The divergent frame rates used with the different systems had the potential to affect our ability to identify the optimal frame within a given acoustic interval. At our lowest frame rate of 21 frames per second, the selected frame could be at most 48 ms from the true frame of interest, which is judged to be sufficient for the present non-dynamic analysis. Although reduced zoom depth may result in fewer pixels available, MCI and NINFL computations are made from overlayed anchors and do not depend on the available number of pixels.

⁴ Because re-elicitations were possible with the young child sample, more than 48 productions were elicited from some child participants. However, some elicitations were later determined to have misaligned ultrasound images or were otherwise unclear (as indicated in the ultrasound measurement section) and were removed from the final data set.

⁵ While broad patterns from this pairwise analysis are reflected in the mixed models, any discrepancies are likely due to there being three comparisons in the mixed models and six comparisons in this analysis. As an example, the mixed models based on MCI found vowels to be less complex than glides/stops/liquids for both adults and children. Based on estimated marginal means, however, vowels were not less complex than glides for adults, and vowels were not less complex than glides/stops for children.

Additional information

Funding

This research was supported by the National Institute on Deafness and Other Communication Disorders of the National Institutes of Health under Grant F31DC018197 (H. Kabakoff, PI), Grant R01DC013668 (D.H. Whalen, PI), and Grant R01DC017476 (T. McAllister, PI). Stimuli from Dawson et al. (2016) were created with support under Grant R01DC002717 (D.H. Whalen, PI). Additional support was provided through an Acoustical Society of American Stetson Scholarship and an American Speech-Language-Hearing Foundation New Century Scholars Doctoral Scholarship.