Objectives/hypothesis: Whereas vibration of the vocal folds has been analyzed and modeled in detail, less attention has been paid to quantifying the slower motions of abduction and adduction. Because these gestures reflect neuromuscular function, cartilage geometry, and joint mobility, objective measurement might improve assessment and treatment of patients with neuromuscular or structural disease.
Study design: Prospective, observational.
Methods: Twenty-one normal adult volunteers performed a repeated "ee-sniff" task at three rates during flexible endoscopy. Distortion in the images was corrected and frame-by-frame analysis of the angle between the true vocal folds (glottic angle) yielded measures of maximum abduction angle (MAA) and mean angular velocities of abduction (MVAB) and adduction (MVAD).
Results: Reliability tests indicated interjudge variability of 3.8 degrees (deg). The mean MAA was 51 deg and ranged from 31 to 77 deg across subjects. The mean MVADs were 229, 309, and 475 deg/s for slow, medium, and fast rates, respectively, and the mean MVABs were 330, 388, and 441 deg/s for slow, medium, and fast rates, respectively. Statistical analysis showed greater influence of gesture rate on adduction than on abduction.
Conclusions: The glottic angle can be reliably measured from flexible endoscopic images and angular velocities of vocal fold abduction and adduction can be determined from analysis of sequential video frames. Accuracy is significantly enhanced by correction of distortion in the images. This approach holds promise for objective characterization of vocal fold function in patients with a variety of disorders.