View Full Version : Re: Linear enveloping of EMG

Jason Harrison
08-30-2000, 08:25 AM
G. Anthony Reina writes:
>We have EMG data from bipolar epimysial electrodes sampled at 500 Hz. A
>bandpass filter from 10 Hz to 1 KHz is applied online along with a
>full-wave rectification and a 5-point moving average (e.g. points 1-5
>become point 1, points 6-10 become point 2, etc.).

This last step is a rather poor way of decimating the signal if you
are going to pass it directly to the 10Hz low pass filter.

>Now we'd like to get the linear envelope of the EMG signal. So a 10 Hz
>low-pass filter is now applied.
>What I'm having trouble seeing is how I have anything left in the
>signal. If I bandpass from 10 Hz to 1 KHz and then low-pass at 10 Hz,
>what is left?

Are you bandpassing, rectifying (non-linear), compressing (moving
average), THEN low pass? Or are you rectifying (non-linear),
bandpassing, compressing (moving average) and then low pass?

If you are doing the first, then you are filtering out noise and the
DC component, then rectifying to compute the "envelope" AND
introducing non-linearities, then you should apply the 10Hz low pass
filter. The second metho should indeed wipe out your signal (assuming
that your filters are symmetric, are of high order, yadda, yadda,


J. Harrison@cs.ubc.ca http://www.cs.ubc.ca/~harrison
Graduate Motto: Free-time with guilt. ftp://ftp.cs.ubc.ca/pub/local/quotes

To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l