In-Ear Pulse Rate Measurement: A Valid Alternative to Heart Rate Derived from Electrocardiography?

Figures

• 

  Figure 1. (a) The Dash Pro with silicone cap in size M; blue arrows show the LEDs for pulse rate detection. (b) Right Dash Pro, worn in the right ear.

• 

  Figure 2. (a) Cosinuss°One; blue arrows show the LEDs for pulse rate detection. (b) Left Cosinuss°One, worn in the left ear.

• 

  Figure 3. Exemplary presentation of pulse and heart rate during the entire test protocol. The dashed frame indicates the change from lying position to cycling. Heart rate of the ECG is depicted as a solid line; the pulse rate of Cosinuss°One is depicted as the dashed line; and the pulse rate of Dash Pro is depicted as the dotted line.

• 

  Figure 4. Exemplary presentation of pulse and heart rate during the 10 min rest measurement in a lying position. The heart rate of ECG is depicted as the solid line; the pulse rate of Cosinuss°One is depicted as the dashed line; and the pulse rate of Dash Pro is depicted as the dotted line.

• 

  Figure 5. Bland–Altman plots using heart/pulse rates ≤90 bpm. Investigated in-ear devices: (a) Cosinuss°One; (b) Dash Pro. Plots indicate differences of the rate values on the y-axis relative to the mean of the two methods (ECG and in-ear measurement) on the x-axis. Limits of agreement (LoA) were calculated as mean ± 1.96 × SD. Biases are depicted as a solid line; LoA are depicted as dashed lines.

• 

  Figure 6. Exemplary presentation of pulse and heart rate (a) during the entire stress protocol; (b) around 100 bpm. The heart rate of ECG is depicted as the solid line; the pulse rate of Cosinuss°One is depicted as the dashed line; and the pulse rate of Dash Pro is depicted as the dotted line.

• 

  Figure 7. Bland–Altman plots using heart/pulse rate ≥100 bpm. Investigated in-ear devices: (a) Cosinuss°One; (b) Dash Pro. Plots indicate differences of the rate values on the y-axis relative to the mean of the two methods (ECG and in-ear measurement) on the x-axis. Limits of agreement (LoA) were calculated as mean ± 1.96 × SD. Biases are depicted as a solid line; LoA are depicted as dashed lines.

• 

  Figure 8. Influence of jaw movements and talking on the PPG signal of both in-ear devices: (a) Cosinuss°One; (b) spectrogram of PPG signal of the Dash Pro.