View Full Version : EMG onset determination summary 2

04-19-2004, 10:24 PM
Dear Biomch-L list serve members:

These are additional replies from my EMG onset query.

Thanks for your contributions,
sujani agraharasamakulam


From: Garry T Allison
Date: Thu, 8 Apr 2004 08:45:09 +0800
To: an_sujani@yahoo.com
Subject: EMG Onsets

The onset of the EMG Signal has to be in the context of the physiological
mechanism you are looking at.
The use of the Shewhart protocols (SD above a baseline) are common but are
(of course) reliant upon the consistent baseline. If there is variance
between the baseline amplitude and variance then the threshold will change
dramatically. The effect of this depends on the rate of increase of the
actual "signal" you are observing.
In the end the relationship between the baseline (noise) and the signal will
relate to the success of the selected algorithm.

I am not convinced the regression analysis (slope and intercept) of the
paper you mentioned (Bui & Hodges) is the best statistic to use. [However
the number of people who refer to this specific manuscript to justify the
3SD above the mean may suggest that I may have the interpretation of the
statistics incorrect & it was published by Electroencephalography and
Neurophysiology....]. So basically I would argue that it is the magnitude of
the error between matched paired that relates to the validity of the METHOD
of determining the EMG onsets. By reporting that there is no difference in
the slope of the curve or the intercept relates to the concordance of the
two methods across a set domain of time (or sample data in this case). These
two elements are not independent and the slope will fit the mean of the data
set...of course a low random error will result in a slope of 1 and intercept
of 0 but these could be sample distribution specific. The RMS of the error
(SEM) in my opinion is less so. Again, if the purpose of testing if 3 SD
above the mean is better than 5SD above the mean then you have to look at
the error of the paired comparisons.

Saying that I have spent a long time trying to find a method of detecting
EMG onsets of trunk muscles - I have come to the conclusion that if the
signal to noise ratio is very high then any algorithm will be fine .. if the
signal to noise ratio is low then one has to be confident to the point where
visual (subjective) assessments have to be made.
Currently we use a protocol different to the Shewhart method and we use an
algorithm to detect the onset and then we have a visual acceptance. We also
use a post test "quality" score for the signal and then repeat these trials
to determine if the visual onset is reliable.

These methods have been reported in:
J Electromyogr Kinesiol. 2003 Jun;13(3):209-16.

Related Articles, Links

Trunk muscle onset detection technique for EMG signals with ECG artefact.
Allison GT.
The Centre for Musculoskeletal Studies, School of Surgery and Pathology, The
University of Western Australia, Level 2 Medical Research Foundation
Building, Rear 50 Murray Street, Perth 6000, Australia. gta@cms.uwa.edu.au

The timing of trunk muscle activation has become an important element in the
understanding of human movement in normal and chronic low back pain
populations. The detection of anticipatory postural adjustment via trunk
muscle onsets from electromyographic (EMG) signals can be problematic due to
baseline noise or electro-cardiac (ECG) artefact. Shewhart protocols or
whole signal analyses may show different degrees of sensitivity under
different conditions.Muscle activity onsets were determined from surface EMG
of seven muscles for five trials before and after fatigue were examined in
four subjects (n=280). The objective of this study was to examine two
detection methods (Shewhart and integrated protocol (IP)) in determining the
onsets of trunk muscles. The variability of the baseline amplitude and the
impact of added Gaussian noise on the detected onsets were used to test for
robustness.The results of this study demonstrate that before and after
fatigue there is a large degree of baseline variance in the trunk muscles
(coefficients of variation between 40-65%) between trials. This could be
normal response to body sway. The IP method was less susceptible to false
onsets (detecting onsets in the baseline window) 3 vs. 51%. The findings
suggest the IP method is robust with large variance in the baseline if the
signal to noise ratio is greater than six.In spite of the robustness of the
algorithm, the findings would suggest that statistical assessments should be
used to target trials for selective visual inspection for subtle trunk
muscle onsets.
Garry T Allison Associate Professor of Physiotherapy
The Centre for Musculoskeletal Studies http://www.cms.uwa.edu.au/
School of Surgery and Pathology, The University of Western Australia.
Level 2 Medical Research Foundation Building
Rear 50 Murray Street
Perth Western Australia 6000.
ph: (618) 9224 0219
Fax (618) 9224 0204


Dear Sujani:

Delsys website has a couple of EMG articles and tutorials you may find it
useful. The website address is: www.delsys.com

devi Bheemappa

Dear Sujani,

I spent a lot of time (some years) in determination of EMD (EMG-on- and
offset) using and comparing all methods and threshold values (with all
concerning advantages and disadvantages) decribed in literature. I think you
may refer to these results and don't need to do this work again.
Unfortunately the whole work is published as a book in german language
(translated: Electromechanical delay of human skeletal muscle ... in 2001).
Only some parts of this work are published in english. You may download
these from the following web-page refering to "Elektromechanische
For the moment we try to transfer all "good" methods to the EMG-part of the
software package of Simi-Motion (Movement analysis www.simi.com).
If you have some further questions or need some more informations please
don't hesitate to contact me.

Yours sincerely

Thomas Jöllenbeck


dear sujani

you can profitably make reference to the paper:

P. Bonato, T. D'Alessio, M. Knaflitz
A statistical method for the measurement of muscle activation intervals from
surface myoelectric signal during gait,
IEEE Trans. on Biomed. Eng., vol. 45, 3, 1998, pp. 287-299,

here you can find a double threshold procedure to determine the onset and
offset of emg signals, with a method which is as far as possible
"objective" that is it does not rely on subjective choices made by the
I understand also that you are undertaking a study on the methods for emg
signal processing in dynamic conditions. In this case, you can also make
reference to:

(for the filtering of electrode movement artifacts)

Conforto S., D¹Alessio T., Pignatelli S.,
Optimal rejection of movement artefacts from myoelectric signals by means of
a wavelet filtering procedure,
Electomyography and Kinesiology, 1999, pp. 47-57

(for a real time and efficient algortihm for the study of fatigue)

Conforto S., D¹Alessio T.,
Real time monitoring of muscular fatigue from dynamic surface myoelectric
signals using a complex covariance approach,
Medical Eng. & Physics, 1999, pp. 225 - 234.

(for an adaptive real time algorithm for the estimation of emg signal

T. D¹Alessio, S. Conforto,
³Extraction of the envelope from surface EMG signals: an adaptive procedure
for dynamic protocols²,
IEEE Engineering in Med. and Biol. Magazine, 2001, pp. 55-61.
do not hesitate to contact me if the need arises

tommaso d'alessio

Tommaso D'Alessio
Dipartimento di Elettronica Applicata
Università degli Studi Roma Tre
Via della Vasca Navale 84
00146 Roma (Italia)
tel. + 39.06.55173266
Fax: + 39.06.5593732
Web site: http://www.dea.uniroma3.it/biolab

To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
Please consider posting your message to the Biomch-L Web-based
Discussion Forum: http://movement-analysis.com/biomch_l