I am doing acquisition of electrodermal activity without filtering, and I have ECG signals associated with my acquisition. Any idea about the characteristics of the filter for elliminate this signal? I prefer a solution implemented on Matlab, but any suggestion could be helpful.
Thank you for the suggestion, it will be very useful in another acquisitions. I am trying to filter ECG (electrocardiogram) information from a signal of electrodermal activity.
Hi Roberto!
Do you mean to obtain the ECG signal from the electrodermal activity or do you want to cancel out this information from the electrodermal activity?
Correct me if I'm wrong, but usually, our vital signal; for instance the heart rate; they have specific frequency that is to say between 1 to 2Hz. So you may need to have a Low Pass Filter if you wish to get this ECG signal.
Hi! I want to cancel the ECG signal, thus the high pass frequency is the best option, but I am not sute if I lose information about electrodermal activity.
HPF is exactly what you need. What kind of electrodermal activity that you're recording now? Since you're afraid your interest signal might be loss if you apply the HPF; then I assume that the electrodermal activity has very low frequency; less than 1 Hz? If so, then the interest signal might be cancelled out as well.
Have you tried to view your signal using fft technique? Matlab has the signal processing toolbox that you can use to straight away 'read' your signal in the power spectrum.
Dear Roberto,
there are some different approaches to this. First of all you will need to know the characteristics of the signal that you want to keep. The significant frequency spectrum of an ECG signal is mostly from 0.5 Hz up to 40 Hz. If your signal of interest overlaps this band, then you cannot use simple frequency filtering technique, unless you can accept loosing information in this band. If you cannot accept loss of information in the ECG band, then you will need some type of ECG template cancellation technique. In the paper attached (Martens 2007), they describe a very good method for cancelling the maternal ECG signal from abdominal recordings in order to isolate the fetal ECG. I think this would be a good solution also to your problem. Let me know if you have any another questions. Costanzo
Si las componentes frecuenciales de la señal electrodérmica se solapan con las de la señal de ECG, una solución puede darse usando Blind Source Separation mediante ICA (independent component analysis) o sparse representation.
As Constanzo well said before me, you need to know the spectral characteristics of the signals you are dealing with. And the ECG has frequency components way above 40 Hz.
Thank yoy for your valuable answers. I will try this solutions in the next week and I will advie to the forum what solution works better.
You need to use few high pass and low pass filters..Take care of noise from AC source ( 60Hz )also because it also adds the noise component in the signal..
Frequency filters are not sufficient in this case as the electrodermal activity and the ECG overlap in their frequency content. You will need some sort of adaptive filtering or source separation algorithm as discussed above.
Dear Roberto ,
if you acquire ECG signal without filter , your signal has to Gauss noise and various noise , such as : 50/60 hz (AC) ...
I think filter tool in matlab is helpful .Butterworth offers do it .
You should use the Butterworth filter, because it will be flat filter the ECG signal without attenuation of the signal amplitude.
You refer to the function of the Butterworth filter in matlab here:
http://faculty.smu.edu/hgray/software/matlabfilterinstructions.pdf
Or examples and samples code :
http://www.egr.msu.edu/classes/ece480/capstone/spring13/group03/documents/SignalProcessingofECGSignalsinMatlab.pdf
Best regard .
Dear roberto,
Try independent component analysis (ICA) to find the source of different signal then alienate them.
Hi everybody. After some weeks of research, the best way for processing the signal was a smoothness algorithm with 1000 samples. Thank you for your suggestions, I hope that this answer sould help you. Good luck!
Hi. I see that this post is very popular, so I want to do an update. My colleague Javier Izquierdo published an article about the study of the beta/delta wave in the study of attention. This parameter concentrates while the person is in awareness and it widespread otherwise. This parameter is robust on noise and shows excellent results in different test.
Article A study of the effects of advanced driver assistance systems...
- Badges
- Science topic