Tweet
Dear all,
Just to put a different perspective I've never heard any consideration of
what might be called the "inertial delay" associated with forces moving
masses. Take the movement of the centre of gravity during gait for example.
It moves vertically in an approximately sinusoidal motion with its maximum
height in mid-stance. This means however that the maximum, acceleration is
in double-support (180 degrees out of phase). Therefore the main muscle
contributors to achieving maximal CM height in mid-stance are those actually
occurring in double-support. This is a much bigger delay (of the order of
250ms) than those being bandied about for EMD. Presumably there must be
similar effects to control movement at specific joints. If you want to
extend (or flex) a joint at a particular point in the gait cycle then you
must apply the muscle force much earlier to achieve this. Obviously the
inertia involved in swing phase phenomena is that of the leg and will lead
for much smaller delays but that for stance phase phenomena is that of the
body and my gut reaction is that those delays will be of similar magnitude
to those for the CM example above.
I've swapped an e-mail with Chris Kirtley before posting this and he draws
attention to his early work
(http://physio.curtin.edu.au:16080/cga/ck/andrews/):
"I was controlling FES with a closed-loop servo and of course I had
horrendous instability problems. Figuring it all out taught me a lot. As you
say, the inertial "delays" caused by the weight of the body segments are
much larger than the muscle delays - I got poles at about 20 Hz for the
electricaly-stimulated muscle and about 1 Hz for the limb dynamics, so of
course the muscle delay was a minor problem."
Any comments?
Richard
Richard Baker
Gait Analysis Service Manager and Director Gait CCRE
Hugh Williamson Gait Laboratory
Murdoch Chidlrens Research Institute
Royal Children's Hospital
Parkville, Victoria 3052, Australia
Tel: +613 9345 5354, Fax: +613 9345 5447
-----Original Message-----
From: * Biomechanics and Movement Science listserver
[mailto:BIOMCH-L@NIC.SURFNET.NL]On Behalf Of David Gabriel
Sent: Wednesday, 15 June 2005 23:40
To: BIOMCH-L@NIC.SURFNET.NL
Subject: [BIOMCH-L] EMD
Greetings All:
In an earlier paper (Gabriel & Boucher, EJAP 1998, 79:37-40), we assumed
that EMD was related to excitation and contraction-coupling and generating
enough tension in the SEC so that force could be transmitted to the tendon
and move the limb. If so, it should be affected by dynamic training which
alters muscle activation . So, we divided the EMG burst into two
components EMD and END. The END portion was from the onset of movement
until the end of the burst. The portions did respond differently. The END
portion of the burst decreased in duration as subjects increased the speed
of limb movement. The duration EMD portion remained "relatively"
stable. We did note that 100 maximal effort contractions decreased its
duration "within" each training session, but were unable to find an across
sessions training effect.
How we filtered the data was an important issue. It affected the absolute
magnitude of the observed results, not the appearance of them. That is,
experimental affects were real, no matter how we filtered the data: i.e.,
band-passed versus linear envelope detection as per Winter (3rd Ed).
In my experience, EMG can be as long as 25-50 ms for isotonic contractions
of the elbow flexors (Gabriel & Boucher, 1998), or as low as 10-15 ms for
isometric contractions of the elbow flexors (Gabriel, Basford, & An, JEK
2001, 11: 123-129). There are figures in both of these papers which show
the differences in EMD for different contraction modalities. Both studies
used a zero phase Butterworth, band-pass filter. The earlier paper also
outline the algorithm for EMG burst onset and termination, which has been
requested many times in this forum.
One last point. I have noticed that EMD values for isotonic contractions
are shorter when using acceleration to determine the onset of movement
versus an event marker triggered by a microswitch; there are additional
'slight' delays associated with the electronics. I also read this is
another paper but can't remember the citation.
Best Wishes,
-d.g.
David A. Gabriel, Ph.D., FACSM
Department of Physical Education and Kinesiology
Brock University
St. Catharines, Ontario, CANADA
L2S 3A1
Phone: 905-688-5550 ext.4362
FAX: 905-688-8364
E-mail: dgabriel@brocku.ca
"I learn from my mistakes. I can repeat them perfectly"
Just to put a different perspective I've never heard any consideration of
what might be called the "inertial delay" associated with forces moving
masses. Take the movement of the centre of gravity during gait for example.
It moves vertically in an approximately sinusoidal motion with its maximum
height in mid-stance. This means however that the maximum, acceleration is
in double-support (180 degrees out of phase). Therefore the main muscle
contributors to achieving maximal CM height in mid-stance are those actually
occurring in double-support. This is a much bigger delay (of the order of
250ms) than those being bandied about for EMD. Presumably there must be
similar effects to control movement at specific joints. If you want to
extend (or flex) a joint at a particular point in the gait cycle then you
must apply the muscle force much earlier to achieve this. Obviously the
inertia involved in swing phase phenomena is that of the leg and will lead
for much smaller delays but that for stance phase phenomena is that of the
body and my gut reaction is that those delays will be of similar magnitude
to those for the CM example above.
I've swapped an e-mail with Chris Kirtley before posting this and he draws
attention to his early work
(http://physio.curtin.edu.au:16080/cga/ck/andrews/):
"I was controlling FES with a closed-loop servo and of course I had
horrendous instability problems. Figuring it all out taught me a lot. As you
say, the inertial "delays" caused by the weight of the body segments are
much larger than the muscle delays - I got poles at about 20 Hz for the
electricaly-stimulated muscle and about 1 Hz for the limb dynamics, so of
course the muscle delay was a minor problem."
Any comments?
Richard
Richard Baker
Gait Analysis Service Manager and Director Gait CCRE
Hugh Williamson Gait Laboratory
Murdoch Chidlrens Research Institute
Royal Children's Hospital
Parkville, Victoria 3052, Australia
Tel: +613 9345 5354, Fax: +613 9345 5447
-----Original Message-----
From: * Biomechanics and Movement Science listserver
[mailto:BIOMCH-L@NIC.SURFNET.NL]On Behalf Of David Gabriel
Sent: Wednesday, 15 June 2005 23:40
To: BIOMCH-L@NIC.SURFNET.NL
Subject: [BIOMCH-L] EMD
Greetings All:
In an earlier paper (Gabriel & Boucher, EJAP 1998, 79:37-40), we assumed
that EMD was related to excitation and contraction-coupling and generating
enough tension in the SEC so that force could be transmitted to the tendon
and move the limb. If so, it should be affected by dynamic training which
alters muscle activation . So, we divided the EMG burst into two
components EMD and END. The END portion was from the onset of movement
until the end of the burst. The portions did respond differently. The END
portion of the burst decreased in duration as subjects increased the speed
of limb movement. The duration EMD portion remained "relatively"
stable. We did note that 100 maximal effort contractions decreased its
duration "within" each training session, but were unable to find an across
sessions training effect.
How we filtered the data was an important issue. It affected the absolute
magnitude of the observed results, not the appearance of them. That is,
experimental affects were real, no matter how we filtered the data: i.e.,
band-passed versus linear envelope detection as per Winter (3rd Ed).
In my experience, EMG can be as long as 25-50 ms for isotonic contractions
of the elbow flexors (Gabriel & Boucher, 1998), or as low as 10-15 ms for
isometric contractions of the elbow flexors (Gabriel, Basford, & An, JEK
2001, 11: 123-129). There are figures in both of these papers which show
the differences in EMD for different contraction modalities. Both studies
used a zero phase Butterworth, band-pass filter. The earlier paper also
outline the algorithm for EMG burst onset and termination, which has been
requested many times in this forum.
One last point. I have noticed that EMD values for isotonic contractions
are shorter when using acceleration to determine the onset of movement
versus an event marker triggered by a microswitch; there are additional
'slight' delays associated with the electronics. I also read this is
another paper but can't remember the citation.
Best Wishes,
-d.g.
David A. Gabriel, Ph.D., FACSM
Department of Physical Education and Kinesiology
Brock University
St. Catharines, Ontario, CANADA
L2S 3A1
Phone: 905-688-5550 ext.4362
FAX: 905-688-8364
E-mail: dgabriel@brocku.ca
"I learn from my mistakes. I can repeat them perfectly"