Analysis of vocal-fold motion from laryngeal image sequences

Analysis of vocal-fold vibration from high speed laryngeal images has been shown to have diagnostic value in the assessment of normal and abnormal voice production. This paper presents a new way to define the boundaries of the vocal-fold opening that is computationally efficient and operates successfully over a range of threshold settings. The area within these boundaries is computed for each image frame to produce the glottal area waveform (GAW) as a function of time. The frequency content of the GAW is analyzed to compare with previous results used for clinical evaluation. In addition, analysis of the glottal area waveform used by others is extended to include new parameters using Fourier descriptors of the boundary contours.