Aim: Electropalatography (EPG) has been used to study objectively the abnormal articulatory patterns associated with cleft palate, however, currently it is difficult to define and quantify abnormal EPG patterns. This study applies three EPG contact indices – the dorsopalatal contact quotient, alveolar contact anteriority index, and alveolar contact centrality index (Recasens, 2004), to identify and quantify abnormal patterns.
Methods: Speakers were four children with articulation disorders associated with cleft palate (three male and 1 female; age from 8-14 years) and four normal children (three male and 1 female; age from 4;01-7;05 years). The PC-based WinEPG system, which records the location and timing of tongue’s contact against the hard palate during speech, was used to record simultaneously the tongue-palate contact pattern and acoustic data during speech production. The speech materials were eight words with target consonants /t k s ʃ/ embedded in the word-initial position; the children were instructed to read aloud these words. The closure period of plosives /t k/ and the frication period of fricatives /s ʃ/ were annotated based on waveform. The three indices were measured at the temporal midpoint of the annotated segment.
Results: In general, these contact indices showed moderate sensitivity and high specificity in classifying abnormal EPG patterns due to articulation errors.
Conclusions: The moderate result was mainly due to the variability in the amount of tongue-palate contact among individuals. It would be of use to further research on the indices on the basis of a larger normal data set.
REFERENCE
Recasens, D. (2004). The effect of syllable position on consonant reduction (evidence from Catalan consonant clusters). Journal of Phonetics, 32, 435-453.