Masticatory Evaluation in Non-contact Measurement of Chewing Movement

Abstract

Decline in masticatory force, which is caused by aging and disease, makes a negative effect on health conditions. As masticatory performance is greatly associated with eating habits, it is important to recognize the chewing states for checking the masticatory function. In this study, the method to classify the characteristics of ingested food, count the number of chewing per one bite, and verify the features of chewing states measuring the chewing movement with a non-contact sensor has been developed. The path and rhythm of chewing movement is evaluated by tracking the feature points on a face measured with an RGB-D camera. The importance of feature quantities extracted from chewing movement was analyzed using statistical approach and machine learning. The results suggested that the chewing states and masticatory function could be evaluated with some parameters such as cycle time, opening distance, and stability of path and rhythm measured with a simplified system.