Methods for measurement of vocal fold vibration and viscoelasticity

Abstract

The overall aim of the thesis was to develop new methods for analysis of vocal fold vibrations and viscoelasticity and to test them in human subjects.

In Study I the onset of vibration and irregular vocal fold vibration was examined with laryngoscopy using a high-speed camera at a frame rate of about 2000 images/sec. A new software called High-Speed Tool Box, HSTB, was developed for automatic analysis. The HSTB can read the sound signal and present the sound exactly synchronized with the image. It can also trace the vocal fold edges during vibration and display the area as a graph. From the high-speed recording it is also possible to make a kymogram, which shows the vibration in one part of image, but for a longer period. Data from one subject with voice tremor and one with diplophonic phonation are presented. When examining the vocal folds with a rigid or flexible endoscope the amplification of the image differs with the distance to the object.

In study II we adapted a laser triangulation method for use in larynx to measure both horizontal distances and vertical movements in a mm-scale. The standard error of measurements in the horizontal plane was between 3-6% and in the vertical plane about 10%.

In study III twenty-seven professional opera singers were examined with laser triangulation to measure the vocal fold sizes related to voice category of the singers. The result showed that the bass group had significantly longer vocal folds than the sopranos and mezzos; also, the males had significantly longer vocal folds than females (p<0.05). Measured values for vocal fold width were significantly larger for males than for females and for the bass group as compared to the other categories (p<0.05).

In study IV a new method for measuring vocal fold viscoelasticity, called Air Pulse Elasticity Measure (APEM), was developed. The method can be used in local anesthesia on human subjects. By blowing controlled air pulses on the vocal folds and measure the resulting mucosal deflections with help of a laser it was possible to calculate a value reflecting vocal fold elasticity. Nine normal vocally healthy subjects were examined with air pulse stimulations on the vocal folds, on the skin (cheek and dorsum of the hand) and on the inside of the lip. The elasticity data showed no differences between the vocal folds, lips or cheeks. The hand data showed significant higher stiffness as compared to the other three measured tissues (p<0.001). The coefficient of variation was about 35% for all measurement points, but in ideal conditions on skin it was 9%. The results indicate that the technique allows automatic quantitative noninvasive vocal fold elasticity measurements on awake subjects even if some methodological development is needed before it can be used clinically.

In study V the APEM was used to measure the elasticity in scarred rabbit vocal folds. Ten scarred New Zealand rabbit vocal folds and 4 normal rabbit folds were measured directly after sacrifice. The elastic data were compared to histological sections from the scarred vocal folds analysed by a pathologist. The results showed significantly lower elasticity (higher stiffness) values for the more scarred vocal folds as compared to samples with minor damage (p=0.03).

In conclusion these new methods are reliable and can be used in practice for analysis of human vocal fold function. However some development is needed before they are clinically useful.