Table 1.
Acronyms and abbreviations.
Fig 1.
Sample ECG and PPG signals with and without effects of respiratory modulation.
(A) unaffected by respiratory modulation. (B) affected by baseline wander. (C) affected by amplitude modulation. (D) affected by frequency modulation.
Fig 2.
Structure of the BiLSTM model.
Table 2.
Performance of BiLSTM NN using various feature vectors for estimating respiratory rate.
Fig 3.
Error Histograms for RR Estimation using RR & RQI features derived from: (A) 20-second PPG & ECG segments. (B) 30-second PPG & ECG segments. (C) 60-second PPG & ECG segments.
Fig 4.
Bland Altman Plots for RR Estimation using RR & RQI features derived from (A) 20-second PPG & ECG segments. (B) 30-second PPG & ECG segments. (C) 60-second PPG & ECG segments.
Fig 5.
Regression Plots for RR Estimation using RR & RQI features derived from: (A) 20-second PPG & ECG segments. (B) 30-second PPG & ECG segments. (C) 60-second PPG & ECG segments.
Table 3.
Comparison to previous works.