2020 Volume 24 Issue 1 Pages 3-13
Purpose: Measurement of lip motion with a compact 3D camera and swallowing dynamics by videofluoroscopy (VF) were performed simultaneously. Synchronization and analysis were performed to clarify the relationship among the distance between angulus oris, swallowing time and swallowing quantity.
Methods: The compact 3D camera was the Microsoft XBox One Kinect Sensor® (Kinect). Development software provided by Microsoft Corporation was used. The face surface of the subject was automatically recognized. A program was created to obtain the three-dimensional coordinates of the both-sides angulus oris and record the distance between the angulus oris. Kinect verified the accuracy and reproducibility of the mannequin distance between the angulus oris under certain conditions.
We examined the most suitable conditions for the measurement of subjects. Based on the verification results, the information obtained by the video and audio synchronization software ELAN was synchronized and analyzed. The subjects were 12 adult men (average age 27.8±1.2 years) with no abnormal swallowing function. Four types of samples were used: barium sulfate suspension 5 mL, 10 mL, 15 mL, and 20 mL. One-way analysis of variance and Pearson’s product-moment correlation coefficient were calculated for the amount of displacement (displacement) and swallowing quantity, swallowing time and swallowing quantity during swallowing. The position of holding the sample in the oral cavity (holding position), lip motion and swallowing quantity were examined by Friedman’ s test.
Results: The optimum conditions (distance 120 cm, rotation angle 10°). The measurement with Kinect had the highest accuracy, with a standard deviation of ±0.52 mm for the distance between the angulus oris and a difference of 0.47 mm between the measurements with Kinect and the actual measurement. There were significant differences and correlations between swallowing and displacement (p＜0.01, r＝ 0.56). There was a significant difference between swallowing quantity and holding position (p＜0.01). However, with respect to the other items, no significant difference was observed.
Conclusions: This study clarified the measurement accuracy and reproducibility of the distance between the angulus oris by Kinect. A system that can measure and synchronize lip motion and swallowing dynamics at the same time was constructed. The swallowing quantity and the displacement were significantly different and correlated, and the swallowing quantity and the holding position were significantly different. On the other hand, there was no significant difference in the other items.
