HRV Measurement to find R-R interval

[infra444]

Hello Friends,

I am trying to measure Heart rate variabilty (HRV) from a contactless capaciteively coupled ECG. I am using Hi speed NI USB -9162 for data acqusition, my data is full of noise..I tried implementing various filters, notch filter, WA detrend, wavelet denoise, but still I am not able to get rid of the noise. I went thru various articles about HRV in labview, but was not succesful with it. 

Though I can get some peaks which are not so clear to measure HRV as you can see from the sample data,  I used multisclae peak detector to find the peaks and then measure the R-R interval between them to find the HRV. I was not successful in attaining this,  I would be really glad if some one can guide me in this regard. I have attached my vi and the sample data here. I am using labview 8.5 version.

Thank you,

abhijit

Message Edited by infra444 on 09-08-2008 03:55 AM

[TBalazs]

Hello Abhijit,

have you already checked the information under: http://zone.ni.com/devzone/cda/epd/p/id/5832?

If you are interested in the source codes of HRV Startup Kit, please send a email to 

zhijun.gu@ni.com or china.info@ni.com

As it is stated at: http://forums.ni.com/ni/board/message?board.id=170&message.id=305721&requireLogin=False 

Best regards,

Blase 

[infra444]

Hello Mr. Blase,

I did go thru that article on HRV, and Implemented the same thing in my Vi, but I was not successful in calculating the heart rate variability from it.

[ZJ.Gu]

Hi,

We don't know how your acquisition system works. Could you please tell us more about your acquisition? We could hardly see the heartbeat from your signals.

The attached VI in LabVIEW 8.5 do resample. filter 50Hz noise, detrend. Hope this will help you.

[infra444]

Hello Mr.Zhijun Gu,

thanks for your reply, I am using Ni USB 9162 daq for data acqusition,it is a four channel +-10 v,16 bit simultaneous analog input with BNC. I have attached my revised vi and sample data, along with a screen shot of my vi,  where I am able to detect the peaks but these peaks are not very constant in amplitude and they are changing continóusaly..so i think i cannot use peak detection to find the heart rate variability. Now all I need to do is  find R-R Interval between the peaks, and plot them as a frequency spectrum.

Is there any other way in which i can find HRV, I dont know if I am correct, I thought of using a mask of standard ECG - pqrst curve on to the final output and then scan them at regular intervals at every 0.5 sec, if the mask and the ecg mach each other then it is a real ecg else its noise. I dont know how this works, it was just my thought.

I would really appreciate any kind of help on this.

Message Edited by infra444 on 09-17-2008 04:19 AM

[ZJ.Gu]

Hi,

Your thought is doable. In fact, what you use now "WA Denoise" is something like search standard ECG, because db06 wavelet is very similar to a standard ECG. However, what I see from your signals is that it is very hard to say that your result is right.  You couldn't add some condition such as two heartbeat interval should be larger than 0.5s, because this is only for a normal relaxed body.

If you think the result you got now is okay, you can set a proper width and threshold of your peak detection VI. (You could delete those wrong R peaks using a reasonable interval threshold.) Then use TSA AR or FFT Spectrum to extract the frequency features.

帖子被Zhijun Gu在 09-17-2008 04:42 AM

时编辑过了

[infra444]

Hi, are there any examples for masking on ECG any reference or vi , I could not find it on the forum.

[Steve.J]

I haven't been able to see the actual data collected from the acquisition system, but I'm unclear on exactly how the data is being acquired.  Are you using any kind of amplifier system before the 9162?  The 9162 was not designed to acquire biosignals directly using surface electrodes.  I don't know about these capacitive contactless electrodes you referenced - are they "active" sensors that provide a low-impedance, amplified output ready to be digitized? 

By the way - we have opened a Life Sciences board just for this sort of discussion here:  http://forums.ni.com/ni/board?board.id=LifeScience

Thanks,

Steve