An automatic method to quantify the vibration properties of human vocal folds via videokymography.

The study offers an automatical quantitative method to obtain vibration properties of human vocal folds via videokymography. The presented method is based on image processing, which combines an active contour model with a genetic algorithm to improve detecting precision and processing speed, can accurately extract the vibration wave in videokymograms and quantify the vibration properties in terms of eight typical parameters automatically. To verify the precision of the proposed algorithm, an indirect simulation setup of vocal folds has been performed. The verification result shows that the relative error of the entire simulation system is less than 5%. Applying the method to analyzing hundreds of videokymograms from 12 subjects, the result indicates that the vibration characteristics of vocal folds can be recognized more exactly, and diseases of the vocal folds can be diagnosed quantitatively. Copyright 2003 S. Karger AG, Basel