Respiratory Rate Estimation

The second fundamental stage of the RR algorithms is 'Respiratory Rate Estimation', in which a RR is estimated from a respiratory signal.

Specify the interchangeable technique(s) for RR Estimation

The techniques to be used for estimation of RRs from respiratory signals must be specified. A brief description of each technique is provided in this publication. Use the following to specify all implemented techniques for use on both ECG and PPG signals:

up.al.options.estimate_rr = {'FTS', 'ARS', 'ARM', 'ARP', 'ARPz', 'ACF', 'WCH', 'PKS', 'ZeX', 'PZX', 'CtO', 'CtA'};

The techniques can be split into frequency- and time-domain techniques. The frequency-domain techniques are as follows:

1. Fourier Transform (FTS , ref)

2. Auto-regressive spectral analysis (ARS, ref)

3. Auto-regressive spectral analysis using the median spectrum for orders 2-20 (ARM, ref)

4. Auto-regressive all-pole modelling (ARP, ref)

5. Auto-regressive all-pole modelling with the lowest frequency pole corresponding to RR (ARPz, ref)

6. Autocorrelation Function (ACF, ref)

7. Welch Periodogram (WCH, ref)

The time-domain techniques are as follows:

1. Peak detection (PKS)

2. Zero-crossing breath detection (ZeX, ref)

3. Combined breath detection using both peak detection and zero-crossings (PZX, ref

4. Count-orig (CtO, ref)

5. Count-adv (CtA, ref)