Toothbrushing monitoring using wrist watch

Daily toothbrushing is essential for maintaining oral health. However, there is very limited technology to monitor the effectiveness of toothbrushing at home. In this paper, a system is built to monitor the brushing quality on all 16 tooth surfaces using a manual toothbrush and an off-Theshelf wrist watch. The toothbrush is modified by attaching small magnets to the handle, so that its orientation and motion can be captured by the magnetic sensor in the wrist watch. The toothbrushing gestures are recognized based on inertial sensing data from the wrist watch. As the acoustic signal collected from the watch is correlated with the motion of toothbrushing stroke, acoustic sensing algorithm is designed to assist in recognition. User-specific toothbrushing order is also utilized to improve the surface recognition. In extensive experiments with 12 users over 3 weeks, our system successfully recognized toothbrushing gestures with an average precision of 85.6%.