Perception and accuracy of an AI-based contactless wellness screening system in dental setting; a cross-sectional study

This study evaluates the reliability and participants' responses to a contactless artificial intelligence (AI) device powered by Remote Photoplethysmography (rPPG) technology for screening vital signs. The contactless AI-based device was compared with automated devices and the manual method as part of its validation for potential healthcare screening applications in dental clinics. A survey was conducted to assess the participants' perception. The Bland-Altman analysis was performed to assess agreement between methods. A total of 499 patients were recruited for this study. There were no significant differences between the contactless AI-based device system and automated or manual measurements for respiratory rate and oxygen saturation (p > 0.05). Significant differences were observed in heart rate, and systolic and diastolic blood pressure compared with the automated device (p < 0.05). Bland-Altman analysis revealed wide limits of agreement for blood pressure (± 30 mmHg for SBP, ± 20 mmHg for DBP) with proportional bias, indicating that BP measurement requires substantial algorithmic refinement before clinical implementation. Respiratory rate and oxygen saturation demonstrated excellent agreement with reference methods. Participants demonstrated high acceptance and positive perceptions of the AI-based system (all mean scores > 4.0) on a 5-point Likert scale. Participants also perceived that people and patients were generally optimistic about the potential of a contactless AI-based device used in dentistry. This study demonstrates that contactless rPPG technology shows validated clinical accuracy for respiratory rate and oxygen saturation monitoring in dental settings. However, blood pressure measurement algorithms require substantial refinement and further validation before clinical deployment. (250 words).