PubMed1999
OBJECTIVE: The analysis of lateral videofluoroscopic images of velar movements during speech is a commonly used tool in the management of the cleft palate patient. This study tests the general hypothesis that measurements of velar movements taken from lateral videofluoroscopic images are accurate and reliable. METHOD: A measurement system was used that allowed for the rapid assessment of velopharyngeal distance, soft palate velocity during the closure cycle, extension of the soft palate at maximum closure, and the angular lift of the soft palate above the plane of the hard palate. Ten recordings of soft palate movement during speech were randomly chosen from lateral X-rays of 27 normal adults. The video recordings were captured by digital frame grabber for subsequent analysis by three operators using a standard PC that was running image-analysis software. The uncertainties associated with the above measurements were analyzed in terms of the errors introduced by the inherent calibration and nonlinearity of the imaging system, the inaccuracy of the patient setup, and the operator-dependent measurement error. RESULTS: For both absolute dimensions and ratiometric measurements, the measurement uncertainties related to the inherent nonlinearity in the imaging system were shown to be less than 2%. Typical patient misalignments as a result of a 10 degree head rotation and a 10-mm translation out of the measurement plane introduced errors of between 2% and 3%. Results showed that the average standard deviation for measurement of gap size was 1.2 mm, extension ratio was 0.11, angular lift was 3.1 degrees, and soft palate velocity was 15.5 mm/second. The intra-class correlation coefficient generally showed a good agreement between operators, typically in the range 0.8 to 0.9. CONCLUSION: Measurements of velopharyngeal distance, extension of the soft palate at maximum closure, and the angular lift of the soft palate above the plane of the hard palate assessed from lateral videofluoroscopic images are reliable and accurate. The soft palate velocity during the closure cycle can also be determined, but clinical interpretations based on this parameter should be constrained by the measurement uncertainties.