MET: Electric Toothbrushing Activity Recognition Based on Magneto-Inductive Sensing

Electric toothbrushes are widely used for home oral care, but many users do not achieve the desired hygiene results due to insufficient brushing coverage or incorrect brushing techniques. The current monitoring systems for electric toothbrushes lack the capability to effectively identify these issues. In this paper, we present a novel electric toothbrush monitoring system, called MET. Compared to existing electric toothbrush monitoring systems, such as Oral-B and Philips, MET achieves four novel functionalities: 1) tracks brushing coverage for all 15 surfaces of teeth, 2) detects different types of incorrect brushing techniques, 3) monitors brushing pressure, and 4) makes just-in-time tooth surface specific brushing mode. This design is inspired by our observation that the rotation of the electric toothbrush motor generates a unique magnetic field, which can serve as a reliable signal for position tracking and operating mode recognition. MET is the first system that tracks both the position and orientation of an unmodified electric motor using magnetic inductive sensing. Experiments with sixteen users show that MET achieves an accuracy of 86.3% in recognizing toothbrushing surfaces, and 91% in recognizing the toothbrush operating modes. Moreover, MET is robust to user location changes and posture variations.