DigiDent: integrated bristle-level sensing and AI for contact-aware toothbrushing assessment

Oral hygiene is fundamental to systemic health, yet toothbrushing, the most common preventive practice, remains largely unguided and unquantified, contributing to the global burden of oral diseases. Current smart toothbrush systems fundamentally fail to capture the critical fine-scale biomechanics at the bristle-tooth interface, which is essential for evaluating clinical adherence and preventing gingival damage. Here we present DigiDent, an AI powered sensing to interpretation system that renders brushing micro-biomechanics computationally interpretable by measuring in-situ contact dynamics at bristle-tooth interface with an embedded E-Bristle sensor. DigiDent fuses bristle-level contact sensing with inertial sensing through a dual-stream neural inference architecture, converting contact-rich brushing behavior into quantitative and interpretable metrics of force, duration, contact-motion pattern, and fine-grained coverage. In validation experiments, DigiDent achieved >99% recognition of motion-contact patterns and 99.84% classification accuracy across 21 brushing subregions. Pilot user-operation verification further demonstrates the feasibility of DigiDent for real-use brushing monitoring and quantitative feedback. These results establish embedded co-deformation sensing and AI-based interpretation as a technical foundation for future personalized toothbrushing assessment and coaching. Oral hygiene is fundamental to systemic health, yet toothbrushing behavior remains largely unquantified because existing monitoring approaches cannot capture the distributed micro-contact dynamics within the bristle field. Here, the authors report DigiDent, a hardware-software co-designed framework combining embedded co-deformation sensing with AI-based analysis to quantify contact-rich brushing behavior. Pilot human studies demonstrate its potential for real-world brushing monitoring, personalized feedback, and habit assessment.