There are some artifacts in my raw EMG data Should I try some filters to remove these artifacts? Could you suggest any solution?
I guess these artifcacts originate from the tissue movement during jump. This will be difficult to eliminate. You could either make a more rigid connection to the skin with extra tape like tegaderm. Or you need another sensor (IMU, gyro, acc etc.) which measures the movement of the body. Then you can eliminate the error if there was a movement. Which sensor you are using?
In future you can use dummy electrode data to record only the motion artifact. For example, place an electrode in the middle of the patella, which should not record muscle activity, if you get comparable motion artifacts you can use a common mode rejection approach to remove the noise. If this does not work, you could also try placing a second electrode close to the location of the recording electrode, but with an insulating barrier between skin and electrode sensors.
Mr. Constantin Wiesener I used Delsys Trigno sensors. It has accelerometer also. I have chechked the change of g force and I know the jumping time. Is it possible to that cropped after jumping time emg datas and analyze the rest? Because artifact happened when the participants were on air.
It's *possible* that a PCA component would load heavily on that segment of your signal and you could then remove that component of you signal from future analysis.
If you are not already pretty familiar with PCA, I'm not sure this should be your first option. Learning what to do and what it means can be time consuming.
It is quiet usual that the system have a delay. If it is bluetooth you will have minimum 25-50 ms. You can check the delay, if you connect a function generator to the electrodes and and ADC input of a data aquisition card and check what the delay is between your measurement and the real signal. It might be frustrating.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators