Project

Doctoral project Kim Bettens

Code
DOCT/005918
Duration
07 March 2012 → 25 May 2016 (Defended)
Doctoral researcher
Research disciplines
  • Medical and health sciences
    • Speech, language and hearing sciences not elsewhere classified
Keywords
hypernasality resonance instrumental diagnostics
 
Project description

For decades, researchers have been searching for the most ideal assessment technique in order to diagnose resonance disorders and to decide on the most apposite treatment. Currently, the presence and degree of resonance disorders is determined by a combination of perceptual judgments and indirect assessment techniques. As speech perception is fundamentally perceptual in nature, perceptual assessments have been traditionally applied to evaluate speech disorders. However, several variables can influence listeners’ perception of resonance which may limit the reliability and validity of perceptual judgments. Consequently, numerous indirect assessment techniques were developed to complement and objectify perceptual assessments. Nevertheless, no indirect technique can yet closely reflect the capabilities of the human ear.

A possible solution to sidestep the limitations of single indirect instrumental assessment techniques is the combination of different variables into a multiparametric index. Following this, Van Lierde et al. (2007) took a first step in creating an instrumental and multiparametric protocol to assess resonance disorders by constructing a ‘Nasality Severity Index’ (NSI). The initial aim of this doctoral thesis was to explore the application of this NSI as a new, multiparametric approach to determine hypernasality in daily clinical practice. To verify the possible influence of personal and environmental variables on the original NSI, the data of 74 Dutch-speaking Flemish children without resonance disorders (mean age (SD): boys 8.3y (2.0), girls 8.4y (2.2); range 4-12y) were analyzed. Based on these data, an influence of age and environmental variables was found, resulting in a large spread of NSI values, even in children without resonance disorders.

Consequently, an adaptation of the original NSI with inclusion of new assessment techniques was aimed, in which the influence of these above-mentioned variables was taken into account. Therefore, additional acoustic techniques to determine hypernasality were explored and included in a statistical analysis to derive a second version of the NSI, namely the Nasality Severity Index version 2.0 (NSI 2.0). Based on the optimal statistical discrimination of 35 children with perceived hypernasality and a control group of 50 children without resonance disorders, a weighted linear combination of three acoustic parameters was established. More specifically, the nasalance value of the vowel /u/ and an oral text passage obtained by the Nasometer and the voice low tone to high tone ratio (VLHR) of a sustained vowel /i/ with a cutoff frequency of 4.47*F0Hz were included. The formula of the adapted NSI yields NSI 2.0 = 13.20 – (0.0824 x nasalance /u/ (%)) – (0.260 x nasalance oral text (%)) – (0.242 x VLHR /i/ 4.47*F0Hz (dB)). With a sensitivity of 92% and a specificity of 100%, using a cutoff score of zero, the NSI 2.0 distinctively discriminates children with hypernasality from children with normal resonance, in which a score below zero indicates the presence of hypernasality.

To implement this new index in daily clinical practice, normative values derived from children and adults without resonance disorders, short-term and long-term test-retest reliability and the relationship between perceptual judgments of hypernasality and the NSI 2.0 and its components were successively determined. Regarding the reference values for the NSI 2.0, no statistically significant influence of gender and age was detected on the NSI 2.0 and its parameters in children. However, significantly lower NSI 2.0 scores were observed in women compared to men, without an effect of age. When the data of children and adults were compared, a significant interaction between gender and age was found for the NSI 2.0 scores, in which adult men showed higher NSI 2.0 scores compared to adult women and children. Based on these study outcomes, separate reference values for the NSI 2.0 and its parameters were established for children, adult men and adult women. With an intraclass correlation coefficient (ICC) of 0.77 in children and 0.84 in adults, NSI 2.0 scores of consecutive measurements are reliable, in which a difference of 2.68 in children and 2.82 in adults can be considered as a genuine change. Additionally, a significant correlation was withheld between the perceptual judgment of hypernasality and the NSI 2.0 scores, in which a more negative NSI 2.0 score indicates the presence of more severe hypernasality. Finally, the NSI 2.0 was applied to objectify the short-term effectiveness of short, intensive speech therapy on the resonance of patients with a history of cleft palate in addition to perceptual assessments.

Considering that the NSI 2.0 can only provide information about hypernasality and the influence of audible nasal airflow on the NSI 2.0 scores is yet unclear, the inclusion of perceptual judgments in the assessment of resonance disorders remains necessary. As both assessment procedures are complementary, they can restrain each other’s limitations and may stimulate critical thinking, especially when contradictory results are observed. In the future, additional instrumental correlates of hypernasality based on connected speech could be further explored. In conclusion, the multiparametric NSI 2.0 forms a new, more powerful approach in the assessment of and treatment planning for individuals presenting with hypernasality.

References

Van Lierde, K., Wuyts, F. L., Bonte, K., & Van Cauwenberge, P. (2007). The nasality severity index: An objective measure of hypernasality based on a multiparameter approach - A pilot study. Folia Phoniatrica et Logopaedica, 59(1), 31-38. doi: 10.1159/00096548