PVC and pulse

In PVC: Premature Ventricular Contraction, I posted some ECG recordings of my heart to show the premature ventricular contractions.  But that post brought up two questions for me:

1. Do the PVCs pump blood around, or do they do nothing?
2. How are PVCs affected by exercise?

In this post, I’ll try to answer the first question, by simultaneously recording both a Lead I ECG signal and a fingertip pulse with an optical pulse monitor.  The amplifier for the optical pulse monitor has a 0.35Hz–5.6Hz filter on it, there is some rounding and delay of the pulse waveform, but I was too lazy to build a new log-transimpedance amplifier without that filter.

I recorded for about 209 seconds, and I got 119 spikes and 56 PVCs.  The overall pulse rate was 50.0bpm, but the sinus rhythm was only about 33.9bpm with 15.8 PVCs/min.  That is a very high rate of PVCs.

I wrote a program that bandpass-filtered  (0.2–50Hz, 4th-order Bessel filter, both forwards and backwards in time) the signal and then identified where all the R spikes were. The program also separated the normal pulses from the PVCs.  This program was specific to my recordings of Lead I—it does not attempt to identify PVCs from an arbitrary ECG!  I also had the program average all the normal waveforms and all the PVC waveforms, lining them up by the detected peak of the R spike.

The solid lines are the ECG signals (in mV) and the dashed lines are the optical pulse signal (in dB, because I was using a log-transimpedance amplifier).

So what can I determine from this plot?

First, the P-spike is missing from the PVCs—this is part of the definition of a premature ventricular contraction, so is hardly surprising.

Second, the T-spike that represents the repolarization wave seems to happen at about the same time for both normal and PVC events, though the shape of the waveforms is different.

Third, my pulse is slow enough and irregular enough that the averaging does not show adjacent ECG signals within this 1.4-second window, but the optical pulse signal shows that the PVC is happening sooner relative to the recovery from the previous pressure pulse than a normal beat. (Again what we would expect from a premature contraction.)

The optical pulse at the fingertip had about the same delay from the R spike for both PVCs and normal pulses—the difference may be due to my defining the R spike as happening at the positive peak, when a more accurate time for the R spike on the PVC would be slightly later.  The PVC pulses were definitely weaker at the fingertips than the normal pulses, but were still fairly strong, so the premature ventricular contractions are pumping blood, but not as much as a normal heartbeat.

In a subsequent post, I’ll try looking at what exercise does to the PVCs.