Table 1.
Participants’ characteristics.
BMI: body mass index.
Fig 1.
Total duration is 44–60 minutes due to variable recovery times.
Fig 2.
(a) ECG leads and Polar OH1 arm sensor placement on the participant. (b) The ECG leads connected to the g.Nautilus device. RA and LL are connected to two detachable electrodes, while LA is connected to the reference (REF) channel of the g.Nautilus. (c) The Polar OH1 sensor on the temple is secured by a sweatband. RA: Right arm electrode, LA: Left arm electrode, LL: Lower left chest electrode.
Fig 3.
Pipeline of deriving the criterion HR from ECG.
ECGHR,i denotes the HR derived at the ith iteration.
Table 2.
Descriptive statistics of criterion measure and the three Polar sensor measurements.
Fig 4.
Aggregate data analysis results.
Bland-Altman plots indicating mean bias scores and 95% limits of agreement (LoA) for the three sensor locations; (a) forearm (b) upper arm and (c) temple.
Table 3.
Results for the aggregated data at different Polar sensor locations (n = 25858), p < 0.0001*.
Table 4.
Data for different HR ranges with respect to the mean HR derived from ECG.
Table 5.
Descriptive statistics and the validation metrics for the entire protocol.
All data other than ICC given in bpm.