I looked at phonocardiography (PCG) simultaneously with ECG and impedance cardiography (ICG) as a method of verifying systolic time intervals.  I assume that you are familiar with the ICG dZ/dt waveform with the vertical axis for amplitude and the horizontal axis for time.  PCG only provides points on the horizontal axis.  For calculating stroke volume, you need some data from the vertical axis.  Time intervals alone are insufficient for reliably calculating stroke volume.  Concerning pre-ejection period, the issue is the reliability of identifying the time point for aortic valve opening.  For ICG, this is called the B point.  There is not a universally accepted standard method, but in general there are 2 approaches.  The first is the point where the dZ/dt line vertically crosses the baseline, which is the average thoracic base impedance (Zo).  The second method is the point where there is an inflection at the base of the ascending limb of the dZ/dt ejection wave.  I use the second method, which I validated with echocardiography.  You need to look at the specific brand of ICG device to see which method is used.  For PCG, the first heart sound (S1 ) is used, but it is not a clean single spike.  If you look at the reference provided by Maxim Malakhov, you see that S1 can be "noisy" with a number of vertical spikes.  For me, it was not easy to consistently select a specific time point for aortic valve opening.  In a clinical setting or research, I do not use PCG.  I prefer ICG.

I use phonocardiography in the past, many years ago. I think that is a very limited method in this moment, and there are many better medical tools to mesure the stroke volume.

Thank you! Jose Luiz Balthazar Jacob, could you specify what other tools you meant regarding human subjects?

Dear Lukasz. Have you some restriction to use image or invasive methods. About the measure of pre-ejection period, I think that phonocardiogram is a good method.  

We have been running experiments in psychophysiology with participants performing tasks sitting in front of a computer, so non-invasive methods are prefered.

Why don´t you use e.g. transthoracic Echocardiography, MR, PET? 

Although systolic-time-intervals have been intensively investigated last decades they never found there way into clinic. You never will convince any clinician by time intervals e.g. that stroke volume is pathologic.

Their acceptance in the scientific literature is also comparatively low. I have have done some (unpublished) scientific work with time-intervals and my opinion is that validity, reliability and objectivity is not high enough to seriously calculate the results. 

You also can not integrate time-intervals into non-invasive standard monitoring because they are prone to disruption an does not contribute to increase efficiency of monitoring.

If you want to apply the time-intervals for testing the amount of vegetative stimuli in humans why don´t you use adequate monitoring? 

Many years ago, I conducted experiments (unpublished) using ICG to study participants performing various stressful tasks, including the playing of video games, while sitting upright.  In my experience, time intervals were not very useful.  The most interesting observations were changes in the shape of the dZ/dt waveform.  The highest point on the dZ/dt wave during systole is called [dZ/dt]max.  As you know, the  ICG formula for stroke volume includes [dZdt]max.  Typically when a participant was stressed, [dZ/dt]max increased in amplitude and shifted to the left, which indicated a stronger contraction.  Another observation was the change of the shape of the dZ/dt wave during systole from triangular to nontriangular.  Neither of these observations can be obtained with phonocardiography.

If you are interested, I can provide more information about the various dZ/dt waveform abnormalities that I observed.