Fuzzy classifier based ingestive monitor

Abstract

The observation of food intake and ingestive behavior remains an open problem that has significant implications in the study and treatment of obesity and eating disorders. A novel method of fusing a sensor and pattern recognition method was developed to detect periods of food intake based on non-invasive monitoring of chewing. A surface-type EMG electrode was used to capture the movement of the lower jaw from volunteers during periods of quiet sitting, and food consumption. These signals were processed to extract the most relevant features, identifying from 4 to 10 features most critical for classifying the type of food consumed. Fuzzy classifiers were trained to create food intake, detection models. The simplicity of the sensor may result in a less intrusive and simpler way to detect food intake. The proposed system is implemented using LabVIEW. The proposed methodology could lead to the development of a wearable sensor system to assess eating behaviors of individuals and also to calculate the quantity of food intake.