SNR for ECG signal

Good answer but almost 3 years too late
M

On Monday, August 8, 2011 at 2:39:26 PM UTC-4, Vladimir Vassilevsky wrote:
> Tim Wescott wrote:
> 
> 
> > Because an ECG signal is either non-stationary or non-Gaussian, or 
> > perhaps both, it's hard to express "SNR".  Is this the measure that 
> > clinicians use to rate the quality of their ECG signals?  If not, what 
> > is?  If there is an alternative that's accepted, why not use it?
> 
> There is aproximately a zillion of ECG analysers of all sorts around. A 
> megatonn of paperwork is written about it. Another ECG processor is just 
> what the world needs.
> 
> Medics are interested in the so-called "QRS complex" parameters 
> calculated from the ECG data. The quality of ECG is the accuracy of the 
> QRS measurement.
> 
> How do I know about this? Because I wrote a math part for a [zillion-1] 
> ECG processor :-)
> 
> Vladimir Vassilevsky
> 
> DSP and Mixed Signal Design Consultant
> 
> http://www.abvolt.com

Hi Vladimir,
            Is Kurtosis of the whole recorded ECG segment a good measure of accuracy of QRS measurement..? 

Thanks,
Naveen 

On Saturday, May 3, 2014 at 5:59:52 PM UTC-4, wdave...@googlemail.com wrote:
> On Thursday, 11 August 2011 16:40:35 UTC+1, Jerry Avins  wrote:
> > On Aug 11, 3:40&#4294967295;am, "steveu" <steveu@n_o_s_p_a_m.coppice.org> wrote:
> > > >On Aug 8, 11:45=A0am, "Jebarajpeter"
> > > ><jebarajselvapeter.p@n_o_s_p_a_m.jasmin-infotech.com> wrote:
> > > >> Hi all,
> > > >> I am working on baseline wander noise removal from ECG signal. I have
> > > to
> > > >> compare the performance of different algorithms. for comparison I have
> > > to
> > > >> use SNR. My question is how to calculate noise and signal power of an
> > > ECG
> > > >> signal available in the data set. I do not know what kind of noise is
> > > the=
> > > >re
> > > >> in the ECG signal.
> > >
> > > >> I have tried like this:
> > >
> > > >> 1. Filter the ECG signal with a basic filter to remove the baseline
> > > wande=
> > > >r
> > > >> signal. Then this signal is considered as Ideal signal (pure signal).
> > > =A0
> > > >> 2. Then I have applied the devised algorithm to the original ECG
> > > signal.
> > > >> So the output from the algorithm has,
> > > >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0output =3D signal + noise;
> > > >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0noise =3D ideal - output
> > > (usin=
> > > >g the algorithm)
> > > >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0signal =3D output - noise;
> > > >> using this i have calculated the SNR. Is it acceptable...? Please
> > > help!!
> > >
> > > >> Best,
> > > >> Jebaraj S.Peter
> > >
> > > >Before you dig into this, find out why the baseline wanders. Ideally,
> > > >you will remove the cause instead of remedying the effect. In any
> > > >event, you will know what you are contending with, always a good
> > > >approach to a problem.
> > >
> > > >Jerry
> > >
> > > You want him to paralyse the patient? :-) ECG traces wander due to muscle
> > > noise - i.e. the muscles other than the heart's.
> > >
> > > Steve
> > 
> > See? It's always good to know what one is contending with. :-)
> > 
> > Jerry
> > --
> > Engineering is the art of making what you want from things you can get.
> 
> Measure your noise in the in ST segmentof the ECG. This is stationary for approx 50mS. So take a window of 30mS in the middle of the S-T and add up all the data samples and find the standard deviation.
> 
> Measure the signal across the QRS postion of the ECG...approx 120mS. Do the same with the data points.
> 
> Divide the signal Std by the Noise Std.
> 
> Job done.

What about the samples in the PQ section? and also samples that are not in the PQRST segment? That is not considered as noise?