Tweet
In reading a few resources for determining the linear envelope of an EMG signal I realized that there is a large range (25-250 ms) used when the "moving window" is established. Winter's 4th edition (biomechanics and motor control of human movement) does not really address this issue explicitly. There is a little more information about setting the moving window in Mereletti and Parker's Electromyograhy (IEEE Press Series in Biomedical Engineering) and Craik and Oatis's, Gait Analysis, Theory and Application, Ch 20. It seems to me that rationales for size of the moving window is scant. I am wondering if anyone can provide a rationale and additional references.
-
-
Re: Linear Envelope
To summarize what I've said previously, in an attempt to generate some more interest in this post; my experience has been that everyone has their own pet theories about the "right" window size - over the years I've seen everything from 50ms to 250ms or longer periods used but I couldn't quote a specific reference for any number. In working with people setting up labs with clinical software over the years, most of the clinicians seem to have a number that they like, but the number is based on past experience, not any specific paper or research.Originally posted by galderink94 View Post
I think that a lot depends on what was happening to generate the EMG - gait reports tend to use longer windows, while folks working on an exercise machine (with shorter activity periods) tend to like smaller windows. As for the literature, my impression is that most of the time you're lucky if the author even tells you the window size and signal bandwidth, let alone links it to a reference - this goes with my theory that people prefer "pretty" and select averaging periods simply because "that's what we've always used." -
Re: Linear Envelope
In part, the answer may be that it depends what you want to do with the signal after your low-pass filter.
I am typically interested in the "on-off" timing patterns and the signal contained during the "on" pattern. Therefore, I do not favor longer windows which distort the onset of EMG activity. As Ed stated previously, this longer window can make the data look "pretty", but I will take accuracy and precision over beauty.
Per ISEK guidelines: "Smoothing the rectified signal with a low pass filter of a given time constant (10-250 ms) is often described as "smoothing with a low pass filter with a time constant of x ms". Time constants higher than 25-30 ms introduce detectable delays and should be used only when interest is on the mean amplitude (moving weighted average) and not on any timing relationship with other events. "
Of course, I also don't favor using the linear envelope in general applications, so I am probably in the minority opinion.Comment
Comment