Estimating caloric intake in bedridden hospital patients with audio and neck-worn sensors

摘要: We present an approach for estimating calorie intake given a limited number of foods provided to patients in in-bed setting. Data collected from proximity sensor, inertial measurement unit, ambient light, and audio sensor placed around the neck are used classify food-type consumed by second using random forest classifier. A multiple linear regression model is then developed each map second-level features calories per second. conducted user study patient simulated lab setting, where 10 participants were asked eat while sitting on bed. user-independent analysis demonstrated detection at 97.2% F1-Score, average Absolute Error 3.0 kCal food-type. Our system shows promise distinguishing food items predicting bedridden participant setting set items.