Automatic estimation of respiratory rate from pulse transit time in normal subjects at rest

Respiratory rate (RR) can often be extracted from electrocardiogram (ECG) or photoplethysmogram (PPG) by analyzing the variations of certain morphological features or the peak-to-peak time interval between them. In this paper, we use a novel method for estimating the RR from the sequence of the pulse transit time (PTT) from radial to the central aorta. To validate this algorithm, in-situ experiments were conducted on ten healthy subjects. For each subject, an ECG, a PPG and a piezoelectric sensor-based reference respiratory signal were collected. The PTT-derived RR was evaluated by comparing with the reference RR. The results showed that the estimation error was 1.44±0.59 breaths per minute (bpm). We also compared the accuracy of the RR derived using this method with that derived from the morphological feature and peak-to-peak time interval of ECG/PPG. We found that the standard deviation (SD) of the PTT-derived results was less than other methods´ SDs. The results demonstrated that the proposed method can be used to estimate the RR in ubiquitous physiological monitoring.