At 01:10, EMG actually records the voltage caused by the depolarisation of the muscle cells, not the action potential before this happens. And at 10:30, rectifying doesn't give you an "average" signal, it gives the absolute value of the signal - i.e., any negative deflections are made positive. And at 15:55, for half-wave rectification, you can just add the absolute signal to the original signal and then divide by 2. And at 18:00, you don't have to rectify a signal before calculating RMS - the process of squaring each value removes the sign. But you don't even seem to be calculating the RMS - you seem to be rectifying the signal then applying a Butterworth low-pass filter twice (filtfilt applies it once in each direction, which is a good idea). Note that when you apply a filter twice, the cutoff changes: the correction factor is 0.802 for a Butterworth applied twice.
25:19 I thought 50Hz-60Hz is an important frequency for EMG signal ? according to a paper I've read. Then isn't by band-stop that frequency, you lost a whole lot of information about the sEMG signal ?
@@LeSandWhich So basically, EMG has a broad spectrum and removing just 50 Hz or 60 Hz isn't going to affect things much, whereas power line interference can easily trick a system into thinking there is a muscle contraction when there isn't. The method described here doesn't analyse the frequency components of the signal anyway, basically doing just a low-pass of the envelope of the rectified signal. Even for diagnostic ECG (heart muscle), the power line interference is normally removed. You can use a steep notch filter that hardly has any effect on other frequencies. You may notice a difference if you took out the whole band from 45 to 65 Hz with a band-stop filter for example, but a notch filter at the right frequency is a good idea. There are various errors in the presentation as I described above. In particular, he confuses the absolute value and the RMS value.
I'm on an old emg system/software which doesn't seem to have an option to normalize data? Is that newer? Can I just amplify the signal and filter instead? Is that what you were doing? Keeping my fingers crossed for a response 🤞
Nice video. Hey, what if your ground reference is much closer to the other 2 electrodes... say, like in the case of the Myo Armband. What's the effect and how should one handle that? thanks for your help!
TopHatRunner Airsoft I'm currently watching the video, and I'm disappointed no on laughed at your observation of the inadequete scotch tape on the surface emg
Hello, now I am working on this technique in my study, could you please send me the powerpoint? I cannot find the PPT in the link given in the comment area. Thx.
Sir...we are final year student.we are doing final year project related to this topic.so we need simulation part.can i get help on that procedure sir..
![The moment we stopped understanding AI [AlexNet]](/img/n.gif)
At 01:10, EMG actually records the voltage caused by the depolarisation of the muscle cells, not the action potential before this happens. And at 10:30, rectifying doesn't give you an "average" signal, it gives the absolute value of the signal - i.e., any negative deflections are made positive. And at 15:55, for half-wave rectification, you can just add the absolute signal to the original signal and then divide by 2. And at 18:00, you don't have to rectify a signal before calculating RMS - the process of squaring each value removes the sign. But you don't even seem to be calculating the RMS - you seem to be rectifying the signal then applying a Butterworth low-pass filter twice (filtfilt applies it once in each direction, which is a good idea). Note that when you apply a filter twice, the cutoff changes: the correction factor is 0.802 for a Butterworth applied twice.
25:19 I thought 50Hz-60Hz is an important frequency for EMG signal ? according to a paper I've read. Then isn't by band-stop that frequency, you lost a whole lot of information about the sEMG signal ?
@@LeSandWhich So basically, EMG has a broad spectrum and removing just 50 Hz or 60 Hz isn't going to affect things much, whereas power line interference can easily trick a system into thinking there is a muscle contraction when there isn't. The method described here doesn't analyse the frequency components of the signal anyway, basically doing just a low-pass of the envelope of the rectified signal. Even for diagnostic ECG (heart muscle), the power line interference is normally removed. You can use a steep notch filter that hardly has any effect on other frequencies. You may notice a difference if you took out the whole band from 45 to 65 Hz with a band-stop filter for example, but a notch filter at the right frequency is a good idea. There are various errors in the presentation as I described above. In particular, he confuses the absolute value and the RMS value.
Thanks Mathil, very educative
I'm on an old emg system/software which doesn't seem to have an option to normalize data? Is that newer? Can I just amplify the signal and filter instead? Is that what you were doing? Keeping my fingers crossed for a response 🤞
You just saved my assignment. Do you have for ECG and His Bundle. You are an amazing lecturer
explains the mechanism of EMG but this is not a clinical application video. Great explanation
Can u give me these slides......
Nice video. Hey, what if your ground reference is much closer to the other 2 electrodes... say, like in the case of the Myo Armband. What's the effect and how should one handle that? thanks for your help!
where can i find this powerpoint?
I tried so hard not to focus on ur face and wondering about ur hair routine to pass my final project discussion in 2 days 😂😂😂
Cool
I would offer a more substantial comment if I had any knowledge in this subject matter
Haha, thanks. I didn't know much on the subject 3 days before I gave the presentation either.
TopHatRunner Airsoft I'm currently watching the video, and I'm disappointed no on laughed at your observation of the inadequete scotch tape on the surface emg
I found it quite funny tho XD
Hello, now I am working on this technique in my study, could you please send me the powerpoint? I cannot find the PPT in the link given in the comment area. Thx.
what data will be input into the mat-lab software? the Arduino ADC?
22:40 what do you mean by axis of the heart?
Interested in doing some work in EMG signal processing? Rob.
Sounds very interesting, but I'm already employed currently
hello i'm working for this topic for my graduating project. can you please send me your thesis document
Hi! Did you find it?
Sir...we are final year student.we are doing final year project related to this topic.so we need simulation part.can i get help on that procedure sir..
We are also doing a project like this
Hi sir, I am doing my final project ( prosthetic arm ) . Can u send me your document so that it may help me . Also the codes
can I get the circuit diagram.
Handsom Airsofter talks stuff to vanila ppl. XD
Sniff....sniff...sniff If you've got a cold don't put the mic next to you...SNIFF SNIFF
Thumbs down