X-Sender: delsys@
Mime-Version: 1.0
Approved-By: Delsys
Newsgroups: bit.listserv.biomch-l
Date: Fri, 12 Feb 1999 10:14:01 -0500
Reply-To: Delsys
Sender: Biomechanics and Movement Science listserver
From: Delsys
Subject: Re: EMG Fundamentals
Comments: To: Trudi Rash

Having recently read an article on the fundamentals of EMG, which was
advertised on this list server (at www.gcmas.org), I feel obligated
clarify a couple of statements which were made:

..."This means in general, one should be collecting in a range from 0 Hz to
600 Hz for surface electrodes and 0 Hz to 1,000 Hz for fine wire
electrodes. Using the Nyquest Theorem, this means that one must sample at a
minimum of 1,200 Hz for surface electrodes and 2,000 Hz for fine wire
electrodes in order to assure capturing the entire signal. Once the signals
have been recorded, then one could use a 10-15 Hz high-pass filter to
eliminate the movement artifacts and a 500 or 1,000 Hz (surface or fine
wire electrode respectively) low-pass filter as an anti-aliasing filter."...

While I don't have a problem with the bandwidths stated, I believe it is
much more advantageous to filter the signal *before* sampling it. This is
particularly important for the antialiasing filter. By definition, the
antialiasing filter is used to prevent the sampling of frequencies in the
signal that are higher than half the sampling frequency, as these
frequencies will be misrepresented if sampled. Any filtering applied after
the sampling process is technically not an antialiasing filter. Filtering
from 600 Hz to 500 Hz once the signal is sampled presents no obvious

I suppose you can argue that no signals exist above 600 Hz and 1200 Hz for
surface and wire EMG, and therefore there is no need for a front end
filter. I don't believe this is a correct argument. Amplifier noise is a
function of bandwidth. By filtering the raw EMG signal before sampling, a
large portion of this noise can be eliminated. This is particularly
important when dealing with low-level EMG signals.

I think it is also advantageous to high pass filter the EMG signals before
sampling. It is possible for motion artifact to be orders of magnitude
larger than the EMG signal. These spike-like disturbances can easily
saturate the amplifiers and exceed the range of the A/D card. Saturated
amplifiers require time to settle, sometimes much longer than the duration
of the artifact. This settling time is dead time as far as EMG signals are
concerned. No amount of digital filtering will restore EMG signals in
these situations. By filtering the artifacts before sampling, a large
portion of these spikes and their associated problems, can be eliminated
from the system.

I understand that filtering signals before sampling complicates that
acquisition process, as digital filters are more convenient to implement
than analog filters. However, I think it is important for people to take
the time to thoroughly understand these issues, and to carefully consider
the pros and cons of both approaches to filtering before setting up an EMG
acquisition system. It may make an important difference.

Gianluca De Luca
Research Engineer


Approved-By: rsdrew@POLYU.EDU.HK
X-Sender: runtech@popd.ix.netcom.com (Unverified)
Mime-Version: 1.0
Newsgroups: bit.listserv.biomch-l
Date: Fri, 12 Feb 1999 15:07:28 -0800
Reply-To: RUN Technologies
Sender: Biomechanics and Movement Science listserver
From: RUN Technologies
Subject: EMG filtering
Comments: cc: ULKYVM.LOUISVILLE.EDUdelsys@delsys.com

Hello to all,

Since Dr. Rash suggested opening up the issue of EMG filtering to
discussion, allow me to add my two cents.

In my opinion, both Dr. Rash and Mr. DeLuca are correct to a certain
extent. Dr. Rash stated, "I say collect 1st, then filter as it is often
difficult to detect artifact in the signal when it is raw, and almost
impossible if you do any manipulation of the data on the front end." I
agree with this statement with one important proviso: aliasing error cannot
be eliminated after the fact. On this point I wholeheartedly agree with
Mr. DeLuca. I want to make sure I am emphatic on this point. Eliminating
aliasing error on the front end should not be considered an option, it's a
necessity. Aliasing occurs when a high frequency signal is inadequately
sampled (i.e., sampled at less than the Nyquist frequency). Inadequate
sampling will not make the higher frequencies go away. Rather, it will
cause them to appear as lower frequencies -- frequencies that are very
likely to be within the spectrum of the muscle being studied. Frequencies
outside the Nyquist range can occur for a number of reasons, such as
capacitance artifacts in the electrodes or lead wires, as well as
electromagnetic interference in the atmosphere. If these sources are not
effectively addressed they can obscure the true signal. And there is
simply nothing you can do about it after the fact.

In all other respects I agree with Dr. Rash. Different muscles have
different fiber compositions, and therefore different frequency spectra. A
hardware device that filters too tightly may be inappropriate for certain
situations. I have heard recommendations that frequencies below 50 Hz can
be effectively ignored(!). Well, that may be approximately true if one is
talking about the large muscles of the leg, or muscles mostly composed of
fast fibers. But that is definitely not true of other muscles. In my
experience, there are situations where even 20-25 Hz can be too high a
cutoff. Another important point is that the hardware filters employed in
many EMG instruments possess broad rolloff characteristics -- usually no
more than about 5-6 db/octave. This can allow high amplitude frequencies
beyond the recording instrument's intended pass-through range to "bleed
through". Using software one can construct extremely efficient filters with
exceptional rolloff characteristics, making them much more effective.
Moreover, the cutoff frequencies of software filters are easily adjustable
as well. For these reasons, my advice always is, use your hardware to get
rid of aliasing, then let your software take care of the rest.

Thank you all for your time.


Rick Lambert

Richard W. Lambert
Dir. of Product Development & Marketing
RUN Technologies
25622 Rolling Hills Road
Laguna Hills, CA 92653 USA
Phone/Fax: (949) 348-1234
email: Rick@runtech.com
web site: http://www.runtech.com



X-Accept-Language: en-US
MIME-Version: 1.0
Approved-By: "Scott Tashman, Ph.D."
Newsgroups: bit.listserv.biomch-l
Date: Mon, 15 Feb 1999 10:17:15 -0500
Reply-To: tashman@bjc.hfh.edu
Sender: Biomechanics and Movement Science listserver
From: "Scott Tashman, Ph.D."
Organization: Henry Ford Health Sciences Center
Subject: Re: EMG Filtering

I agree strongly with Gianluca De Luca's and Rick Lambert's position on
anti-aliasing filters for EMG. There are many possibililities for
high-frequency contamination of EMG data. For example, our lab is not far from
a 50kW broadcast station, and we sometimes see 90MHz noise in raw data. Many
(if not most) computers introduce high-frequency noise into acquired signals,
especially if the acquisition card is located within the computer chassis. When
data contaminated with noise above half the sampling frequency has been
acquired, the noise is aliased down to lower frequencies and cannot be removed.
This noise often ends up looking much like low-level EMG in the acquired
signals, os you may have it and never know about it. There are only three ways
that I know of to avoid aliasing errors: collect perfect, noise-free data, use
unreasonably high sample rates (in the case of our radio signal, 180MHz), or use
front-end analog low-pass filters. The first two options are not realistic
under most circumstances.

As for high-pass filters for motion artifact, there is more room for
discussion. There is certainly no advantage to collecting EMG data down to DC
(0 Hz); the question is, what frequency range is appropriate? Large,
low-frequency artifacts often occur in surface and fine wire EMG. If amplifier
gains are set low enough to avoid saturation and the associated instrumentation
problems Gianluca described, then signal-to-noise ratio is compromised. But the
intended use of the EMG must also be considered. If the data are to be used for
identifying muscle timing and relative activity levels, loss of the small amount
of very low frequency EMG signal which may be occur under some circumstances
will not affect data analysis. For most EMG users, I believe that analog
high-pass filters (cutoff 30Hz or so) improves the reliability and quality of
the EMG recording, without loss of any useful information. This is particularly
true for movements involving impact (e.g. jumping), where motion artifacts are
greatest. However, if one is particularly interested in the frequency content
of the signal (e.g. fatigue studies), and/or looking at relatively slow
movements, then a valid argument could be made for avoiding analog high-pass

If anyone really wants to know about the frequency characteristics of a
particular muscle, they can acquire data during an isometric contraction (to
avoid motion artifact) using a high sampling rate and no high-pass filter, and
then use the FFT to look at the power spectrum of the signal. This analysis is
now relatively routine and can be done in most software being used for EMG, and
is a good exercise for anyone who works with EMG to increase understanding of
the nature of the signal. Then you can make your own decision about how much is
lost for a particular filter cutoff frequency.

The argument about storing "raw" vs. "processed" EMG has been around for quite a
while, but has usually been in reference to envelope-processing schemes which
greatly reduce information content and bandwidth. I am a strong advocate for
storing "raw" EMG, and performing any additional manipulation in the digital
domain. However, I think the use of analog pre-filtering is well accepted in
our field and is usually beneficial.

One final comment: filtering should never be used as a substitute for good lab
technique. No amount of filtering will overcome inadequate skin preparation,
poor attachment of electrodes and leads, use of long, unshielded cables, etc.
If your data is heavily contaminated with noise, the major sources of the noise
should be determined and eliminated if possible. Then, intelligent decisions
about filtering can be made.
Scott Tashman, Ph.D.

Head, Motion Analysis Section Assistant Professor
Bone and Joint Center Department of Orthopaedics
Henry Ford Hospital School of Medicine
2799 W. Grand Blvd. Case Western Reserve University
Detroit, MI 48202

Voice: (313) 916-8680 or 916-7572
FAX: (313) 916-8812 or 916-8064
Internet: tashman@bjc.hfh.edu

Return to PE 506 page