I have used Kubios with great success. Yet, I purely used it for analysis, not for R wave detection. I used other (Clampfit or custom) software that detected the R wave, and thus RR interval. I was able to manually remove artifacts as well as RR intervals that were more than 2 standard deviations from the mean RR. I exported it as a text file that I loaded into Kubios.

As you have written, it is possible to manually remove or add markers of R waves, thus correcting the RR intervals. This is in my experience the only reliable approach with Kubios.

Regards,

mojca

Cave canem. If you look at the ECG you may find that the odd RR intervals are absolutely legitimate. If you have a fast patient who has a pause, then the interval which is anyway 2x longer than the standard RR of this patient (hence probably you would call it odd) it may still be shorter than 2500 ms which is a standard upper range for normal RR intervals. So it matters what is the question you ask. Some tools require strict stationarity, then you remove. Sometimes you specifically look at oddities then you don't remove.

Concerning the technology +1 for David S. Auerbach, I would use a custom made Python script, it is vary good in text file processing.

Concerning the correction.There are two levels. The first is that if you are suspicious that your R detection is not perfect, you need a tool which requires ECG to run. I don't know any tools that are ready to use, however labelling and tagging is something you actually do quite often with biomed signal, so maybe some generic solutions exist (GUI is typically the biggest problem). The second level is if you believe in your R detection and only want to have nicer signal, then as I wrote it makes more or less sense, depending on the experimental protocol. In this case  you don't need ECG.

I never used kubos but only used labchart. It has features that you want.

Hi, 

I support David option as an easy and fast way to do that. 

Greetings. 

Elena. 

Hi

Are you interested in the hrv time series or the power spectra? If its spectra you are seeking there are automatic methods which allow you to bypass aberrant beat removals.

Ed

.

Yes, it's possible.

. Check the Kubios manual - Section 4.4.2

     regards    Ivan Domuschiev

Hi, u have not mentioned any thing regarding how u record HRV I mean acquiring software. Once u do the recording try converting and extracting the RR interval in ASCII format into a notepad file. Open the saved note pad file u can actually go through individual RR value. In this u can see very high or very low RR values which are usually caused by ectopics. Remove them. Close the file and read it using Kubios. This is just a crude way. 

As far as my experience concern, if you have many artifactcs, your time series analysis of HR become not reliable in any case and you shoud reject your ECG samples. In fact, if you delete "manually" the "bursts",  you compromise your analysis because  the lack of data induces anomalus prolongations of RR interval. On the other hand, If you include "manually" (yes you can in most software and I use Thomaseth K ...) more than few RR intervals  you compromise your analysis as well because you insert artificial values... Moeover, I would be sceptic in the use of high band "filters"  as  they can modify the collection of RR sampling... It is a long talk and I hope to be synthetic enough althoug probably not encouraging, I am afraid.

Yes I agree with  Frederico as at some point the data becomes unusable. There are crutches- such as performing splines through the occasional rr gaps. 

Just to summarize shortly the options you have in Kubios:

1) You can manually check and correct (if necessary) the artefacts by editing the R-wave detection markers presented in the ECG data axes. This approach is naturally availble only when you have imported ECG data into Kubios.

2) To correct artefacts automatically, you can use the artefact correction tools and set the correction level such that all clear artefacts are corrected, but normal beat intervals are not modified.

If I have the ECG, then I like to go through with first option, and rely on the second option only if necessary (e.g. if the ECG is so noisy for a short interval that you cannot visually observe the correct time instances for R-waves). 

Hi all,

Apologies for the delayed reply. Thanks for all the help. I now have a much better understanding of what I want to do in Kubios. 

In particular, I was not certain about whether to use automatic or some sort of manual editing so clarification on this has been important but so has the elaboration and further detail. 

In my case I do have ECG data and so then, from my understanding, I will want to focus on manual artifact removal. To do this I would simply remove the R wave detection markers only (rather than using the other two option when right clicking on markers such as add and edit I think) and by doing this the RR interval data will change accordingly so that the artifacts will no longer be present. I should end up with better data which neither includes artifacts nor removes real data points as possible with automatic approach. 

This makes better sense now. Thanks again for all the help all.