Commercial smartwatch for respiratory rate monitoring: Validation against clinical reference methods
Simon Walzel, Filip KonradyAim
The aim of this study was to evaluate the accuracy of respiratory rate (RR) estimation obtained from a commercial smartwatch in comparison with clinical reference methods.
Methods
A total of 58 healthy participants were included in the final analysis. RR was simultaneously measured using a smartwatch (Masimo Watch W1) and reference methods based on spirometry, capnography, and ECG-derived impedance pneumography. Measurements were performed during spontaneous breathing and controlled breathing guided by a visual animation with predefined RR (13, 20, 13, and 6 min -1 ). Accuracy was evaluated using correlation and Bland–Altman analysis and root mean square error (RMSE). Dynamic performance was assessed based on detection delays during transitions between RR.
Results
During spontaneous breathing, the smartwatch demonstrated a non-significant bias (0.47 min -1 ) and an RMSE of 1.92 min -1 . Under controlled breathing, RMSE was higher compared to reference methods (4.37 vs. 3.18–3.56 min -1 ), which decreased after temporal alignment (2.61 vs. 1.44–1.79 min -1 ). The smartwatch exhibited greater variability and a tendency toward slower response during transitions, while still capturing overall RR trends.
Conclusions
The Masimo Watch W1 provides accurate photoplethysmography-based RR estimates during spontaneous breathing in healthy adults, with performance comparable to reference methods. Although reduced accuracy was observed during dynamic RR changes across the evaluated RR range of 6–20 min -1 , the device remained capable of capturing RR trends. These findings support the potential use of smartwatch-based RR monitoring for continuous, unobtrusive trend assessment and screening in similar populations.