Tweet
Dear all,
Dear all,
I agree with Chris that in moving forward to new marker sets we are going to
have to demonstrate clearly the benefits of doing so. The main limitation of
the conventional models is that they are extremely sensitive to marker
placement error (read the studies of Noonan, Gorton 2000 and Gorton 2001 if
you need proof). We've done extensive testing in our lab and very
experienced testers can get acceptable repeatability. Even moderately
experienced testers struggle to get acceptable results though. Intensive
quality assurance in labs is essential.
We've just completed an analysis of a new approach based on kinematic
fitting which I'll be reporting at the International Society of Biomechanics
meeting in Taiwan later this morning
We're concentrating on the pelvis, hip and knee and have markers over the
PSIS and ASIS on the pelvis, two anterior thigh markers (about 1/3 and 2/3)
along the length of the femur and one lateral thigh marker (about 1/2 way).
We then have two markers on the anterior tibial crest and a medial and
lateral malleolar marker. We define hip joint centres and the alignment of
the knee joint axis using functional calibration. We've called the model we
use "Kylie" (because she comes from Melbourne and does the locomotion!).
We've done a repeatability study of 10 normal subjects using two highly
experienced Plug In Gait marker placers comparing repeatability of Plug In
Gait with Kylie. The two crack marker placers do really well but Kylie does
even better with the standard deviation of the inter-session variability
less than 2 degree across all planes for all joints (pelvis, knee and hip -
except the transverse plane of the knee where the error creeps up to about 4
degrees for reasons we haven't been able to identify yet). The Kylie
approach still needs careful placement of markers on the pelvis and ankle
markers but is largely independent of the precise placement of the other
markers and does not need the use of a knee alignment device or similar.
The work so far has just been on healthy kids and we need to validate in our
clinical population as well but we hope that we will have achieved this by
the end of the year and will be able to make further details available for
anyone else who'd like to test the technique.
Richard
Richard Baker PhD CEng CSci
Director Gait CCRE/Gait Analysis Service manager
Murdoch Childrens Research Institute
Royal Childrens Hospital
Parkville 3052, Victoria, 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 Chris Kirtley
Sent: Tuesday, 3 July 2007 11:46 PM
To: BIOMCH-L@NIC.SURFNET.NL
Subject: Re: [BIOMCH-L] Summary of marker sets and gait models
Dear Nancy/others interested,
I was a bit disappointed by the brevity of this discussion, so I thought I'd
throw a few observations in to re-energise it.
It seems to me that the question of which marker set/model is better is not
entirely a scientific one. The Helen Hayes model has the great advantage of
familiarity - it has been in use for over 10 years and it is relatively well
understood. People are naturally reluctant to switch to a new model for
several reasons:
1. They will need to learn and familiarise themselves with its
idiosyncrasies (they are inevitable)
2. They will have the problem of how to compare data collected with two
different models
3. Until a substantial number of labs switch, they will - at least for a
period of time - be somewhat isolated from other labs (e.g. publishing
papers, comparing results, presenting at conferences etc.). Indeed, if the
set does not gain widespread adoption, they will have ended up down a blind
alley.
I'm sure most labs out there are aware of these problems, and that's why
Helen Hayes (the model, not the actress) reigns supreme. Any pretender will
thus, unfortunately, have to be vastly superior in order to compete with the
inertia that HH has developed.
Chris
On 6/29/07, Denniston, Nancy wrote:
>
> I have received a wonderful response from the Biomch-L audience. I have
> asked each of the contributors who had e-mailed directly to me for
> permission to summarize their responses here. I thank them all for their
> insights, reference material and ideas and hope this may continue some
> discussion.
>
>
>
> Nancy Denniston
>
> Center for Gait and Movement Analysis
>
> Denver, Colorado
>
>
>
> ----- Original Message -----
>
> From: "Denniston, Nancy"
>
> To:
>
> Sent: Wednesday, June 27, 2007 12:43 PM
>
> Subject: [BIOMCH-L] Marker sets and gait models
>
>
>
>
>
> We are exploring other marker sets for walking models that we could use in
> lieu of our present use of the Helen Hays set-curious to see what some of
> the other places are using in terms of lower body sets with or without
> multi-segment foot model, and full body sets. Models that have 6 degrees
of
> freedom as well as those that do not .
>
>
>
> Any information on this would be appreciated.
>
> Nancy Denniston
>
>
>
>
>
>
>
>
>
> Dr. Harald Böhm writes:
>
> We are using motion analysis for evaluation and optimization of sport
> equipment. To my experience different movements as well as different
motion
> tracking systems require different marker sets. We are often forced to use
> our own marker set depending on the systems requirements (Vicon, SIMI,
..)
> and the motion to be tracked (golf, gait, skiing, ...) . For the typical
> example of Gait analysis using a Vicon automatic tracking system we would
> use the following steps to define our own marker set:
>
>
>
> 1. It is essential to define the anatomical landmarks based on markers
> (at least during a standing trial) to obtain the joint angles defined by
the
> ISB standardisation e.g. (Grood, E.W., Suntay, W.J., 1983. A joint
> coordinate system for the clinical description of three-dimensional
motions:
> Applications to the knee. Journal of Biomedical Engineering 105, 97-106).
> After a standing trial some anatomical landmarks on the medial side must
be
> removed since they will fall off during gait anyways.
>
>
>
> 2. Vicon requires minimally 3 markers for a rigid segment (foot, shank
> thigh ...) to track the segment automatically during the motion. The
> tracking is improved when the markers of one segment are not moving
relative
> to another. This is best when the markers are attached to a bony surface
> e.g. the tibia. More markers are helpful for the tracking and might
> improve the calculations (Soederkvist, I, Wedin, P.A., 1993. Determining
> the movements of the skeleton using well-configured markers. Journal of
> Biomechanics 26, 1473-1477) or (Journal of Biomechanics 34 (2001) 355±361
> Correcting for deformation in skin-based marker systems Eugene J.
Alexander,
> Thomas P. Andriacchi). The markers of the left and right side body side
must
> be slightly different otherwise the tracking of the actual left and right
> side is not reliable. With this approach 6 DOF for all segments are
> defined.
>
>
>
> 3. The markers must be visible during gait which depends often on the
> number of the cameras you are using. This can be done usually only by
> testing and it is very time consuming. Using a standard marker set from a
> company usually works well for their product and their specified movement.
>
>
>
> Regarding tests of calculations based on marker sets a good collection of
> methods can be found at: (Journal of Biomechanics 39 (2006) 1778-1786
> Determining rigid body transformation parameters from ill-conditioned
> spatial marker co-ordinates A.B. Carman, P.D. Milburn)
>
>
>
> Dr. Harald Böhm
>
> Technical University Munich
>
> Faculty of Sport Science
>
> Department Sport Equipment and Materials
>
>
>
> Dr. Patria Hume writes:
>
> I have recently conducted a sabbatical with Benno Nigg and Gweneth DeVries
> at Uni of Calgary on a LisFranc fracture project. We developed a new
> multi-segment foot model as part of that project. We are currently
> finalising papers on the project and the foot model, soon to be submitted
> for publication and has yet to be reviewed.
>
>
>
> Associate Professor Patria Hume
>
> Director, Institute of Sport and Recreation Research NZ
> Division of Sport and Recreation, Faculty of Health & Environmental
> Sciences
> AUT University
> Auckland, New Zealand
>
>
>
> David Carmines writes:
>
> We use the regular Helen Hayes marker set. For the upper body, we only
> use the C7, T10, CLAV, STRN, SHO, ELB, LWA, LWB, and Fin and not the
> mid-upper arm or mid-lower arm, except where research requires them. Our
> skeletons appear with the above minimal set (plus the 4 head markers for
the
> skull - very important in a skeleton!). That saves us time and provides
> fewer things to go wrong. However, we only use the skeleton for the upper
> body, and I can't confirm that the rotations and center of mass is correct
> without all the arm markers.
>
>
>
> Dave Carmines
>
> Clinical Engineer
>
> Kluge Children's Rehab. Ctr. - UVA
>
>
>
> Alec Black writes
>
> I think the gist of what we are doing is testing other marker sets for
> clinical evaluation using Visual 3D. Those marker sets include a modified
> Helen Hayes, the NIH model and the Oxford foot model.
>
>
>
> Alec Black
>
> Director, Shriners Gait Lab
>
> Sunny Hill Health Centre for Children
>
> Vancouver, B.C.
>
>
>
> Note from Nancy: another site reports using a Cleveland Clinic marker set
> (full or lower body only) with an AI duPont multisegment foot.
>
>
>
>
>
>
>
>
> CONFIDENTIALITY NOTICE: The information contained in this message is
> legally
> privileged and confidential information intended for the use of the
> individual or
> entity named above. If the reader of this message is not the intended
> recipient, or
> the employee or agent responsible to deliver it to the intended recipient,
> you are
> hereby notified that any release, dissemination, distribution, or copying
> of this
> communication is strictly prohibited. If you have received this
> communication in
> error, please notify the author immediately by replying to this message
> and delete
> the original message.
> Thank you
>
> ---------------------------------------------------------------
> Information about BIOMCH-L: http://www.Biomch-L.org
> Archives: http://listserv.surfnet.nl/archives/Biomch-L.html
> ---------------------------------------------------------------
>
--
Dr. Chris Kirtley MB ChB, PhD
from 1 May to 31 July 2007 I am at:
Stiftung Orthopädische Universitätsklinik Heidelberg
Leiter Ganganalyselabor
Anschrift: Schlierbacher Landstr. 200a
69118 Heidelberg
Tel: 49+06221-96 6724
Fax: 49+06221-96 6725
Clinical Gait Analysis: http://www.univie.ac.at/cga
Book:
http://www.amazon.co.uk/exec/obidos/ASIN/0443100098/203-6674734-4427132
Dear all,
I agree with Chris that in moving forward to new marker sets we are going to
have to demonstrate clearly the benefits of doing so. The main limitation of
the conventional models is that they are extremely sensitive to marker
placement error (read the studies of Noonan, Gorton 2000 and Gorton 2001 if
you need proof). We've done extensive testing in our lab and very
experienced testers can get acceptable repeatability. Even moderately
experienced testers struggle to get acceptable results though. Intensive
quality assurance in labs is essential.
We've just completed an analysis of a new approach based on kinematic
fitting which I'll be reporting at the International Society of Biomechanics
meeting in Taiwan later this morning
We're concentrating on the pelvis, hip and knee and have markers over the
PSIS and ASIS on the pelvis, two anterior thigh markers (about 1/3 and 2/3)
along the length of the femur and one lateral thigh marker (about 1/2 way).
We then have two markers on the anterior tibial crest and a medial and
lateral malleolar marker. We define hip joint centres and the alignment of
the knee joint axis using functional calibration. We've called the model we
use "Kylie" (because she comes from Melbourne and does the locomotion!).
We've done a repeatability study of 10 normal subjects using two highly
experienced Plug In Gait marker placers comparing repeatability of Plug In
Gait with Kylie. The two crack marker placers do really well but Kylie does
even better with the standard deviation of the inter-session variability
less than 2 degree across all planes for all joints (pelvis, knee and hip -
except the transverse plane of the knee where the error creeps up to about 4
degrees for reasons we haven't been able to identify yet). The Kylie
approach still needs careful placement of markers on the pelvis and ankle
markers but is largely independent of the precise placement of the other
markers and does not need the use of a knee alignment device or similar.
The work so far has just been on healthy kids and we need to validate in our
clinical population as well but we hope that we will have achieved this by
the end of the year and will be able to make further details available for
anyone else who'd like to test the technique.
Richard
Richard Baker PhD CEng CSci
Director Gait CCRE/Gait Analysis Service manager
Murdoch Childrens Research Institute
Royal Childrens Hospital
Parkville 3052, Victoria, 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 Chris Kirtley
Sent: Tuesday, 3 July 2007 11:46 PM
To: BIOMCH-L@NIC.SURFNET.NL
Subject: Re: [BIOMCH-L] Summary of marker sets and gait models
Dear Nancy/others interested,
I was a bit disappointed by the brevity of this discussion, so I thought I'd
throw a few observations in to re-energise it.
It seems to me that the question of which marker set/model is better is not
entirely a scientific one. The Helen Hayes model has the great advantage of
familiarity - it has been in use for over 10 years and it is relatively well
understood. People are naturally reluctant to switch to a new model for
several reasons:
1. They will need to learn and familiarise themselves with its
idiosyncrasies (they are inevitable)
2. They will have the problem of how to compare data collected with two
different models
3. Until a substantial number of labs switch, they will - at least for a
period of time - be somewhat isolated from other labs (e.g. publishing
papers, comparing results, presenting at conferences etc.). Indeed, if the
set does not gain widespread adoption, they will have ended up down a blind
alley.
I'm sure most labs out there are aware of these problems, and that's why
Helen Hayes (the model, not the actress) reigns supreme. Any pretender will
thus, unfortunately, have to be vastly superior in order to compete with the
inertia that HH has developed.
Chris
On 6/29/07, Denniston, Nancy wrote:
>
> I have received a wonderful response from the Biomch-L audience. I have
> asked each of the contributors who had e-mailed directly to me for
> permission to summarize their responses here. I thank them all for their
> insights, reference material and ideas and hope this may continue some
> discussion.
>
>
>
> Nancy Denniston
>
> Center for Gait and Movement Analysis
>
> Denver, Colorado
>
>
>
> ----- Original Message -----
>
> From: "Denniston, Nancy"
>
> To:
>
> Sent: Wednesday, June 27, 2007 12:43 PM
>
> Subject: [BIOMCH-L] Marker sets and gait models
>
>
>
>
>
> We are exploring other marker sets for walking models that we could use in
> lieu of our present use of the Helen Hays set-curious to see what some of
> the other places are using in terms of lower body sets with or without
> multi-segment foot model, and full body sets. Models that have 6 degrees
of
> freedom as well as those that do not .
>
>
>
> Any information on this would be appreciated.
>
> Nancy Denniston
>
>
>
>
>
>
>
>
>
> Dr. Harald Böhm writes:
>
> We are using motion analysis for evaluation and optimization of sport
> equipment. To my experience different movements as well as different
motion
> tracking systems require different marker sets. We are often forced to use
> our own marker set depending on the systems requirements (Vicon, SIMI,
..)
> and the motion to be tracked (golf, gait, skiing, ...) . For the typical
> example of Gait analysis using a Vicon automatic tracking system we would
> use the following steps to define our own marker set:
>
>
>
> 1. It is essential to define the anatomical landmarks based on markers
> (at least during a standing trial) to obtain the joint angles defined by
the
> ISB standardisation e.g. (Grood, E.W., Suntay, W.J., 1983. A joint
> coordinate system for the clinical description of three-dimensional
motions:
> Applications to the knee. Journal of Biomedical Engineering 105, 97-106).
> After a standing trial some anatomical landmarks on the medial side must
be
> removed since they will fall off during gait anyways.
>
>
>
> 2. Vicon requires minimally 3 markers for a rigid segment (foot, shank
> thigh ...) to track the segment automatically during the motion. The
> tracking is improved when the markers of one segment are not moving
relative
> to another. This is best when the markers are attached to a bony surface
> e.g. the tibia. More markers are helpful for the tracking and might
> improve the calculations (Soederkvist, I, Wedin, P.A., 1993. Determining
> the movements of the skeleton using well-configured markers. Journal of
> Biomechanics 26, 1473-1477) or (Journal of Biomechanics 34 (2001) 355±361
> Correcting for deformation in skin-based marker systems Eugene J.
Alexander,
> Thomas P. Andriacchi). The markers of the left and right side body side
must
> be slightly different otherwise the tracking of the actual left and right
> side is not reliable. With this approach 6 DOF for all segments are
> defined.
>
>
>
> 3. The markers must be visible during gait which depends often on the
> number of the cameras you are using. This can be done usually only by
> testing and it is very time consuming. Using a standard marker set from a
> company usually works well for their product and their specified movement.
>
>
>
> Regarding tests of calculations based on marker sets a good collection of
> methods can be found at: (Journal of Biomechanics 39 (2006) 1778-1786
> Determining rigid body transformation parameters from ill-conditioned
> spatial marker co-ordinates A.B. Carman, P.D. Milburn)
>
>
>
> Dr. Harald Böhm
>
> Technical University Munich
>
> Faculty of Sport Science
>
> Department Sport Equipment and Materials
>
>
>
> Dr. Patria Hume writes:
>
> I have recently conducted a sabbatical with Benno Nigg and Gweneth DeVries
> at Uni of Calgary on a LisFranc fracture project. We developed a new
> multi-segment foot model as part of that project. We are currently
> finalising papers on the project and the foot model, soon to be submitted
> for publication and has yet to be reviewed.
>
>
>
> Associate Professor Patria Hume
>
> Director, Institute of Sport and Recreation Research NZ
> Division of Sport and Recreation, Faculty of Health & Environmental
> Sciences
> AUT University
> Auckland, New Zealand
>
>
>
> David Carmines writes:
>
> We use the regular Helen Hayes marker set. For the upper body, we only
> use the C7, T10, CLAV, STRN, SHO, ELB, LWA, LWB, and Fin and not the
> mid-upper arm or mid-lower arm, except where research requires them. Our
> skeletons appear with the above minimal set (plus the 4 head markers for
the
> skull - very important in a skeleton!). That saves us time and provides
> fewer things to go wrong. However, we only use the skeleton for the upper
> body, and I can't confirm that the rotations and center of mass is correct
> without all the arm markers.
>
>
>
> Dave Carmines
>
> Clinical Engineer
>
> Kluge Children's Rehab. Ctr. - UVA
>
>
>
> Alec Black writes
>
> I think the gist of what we are doing is testing other marker sets for
> clinical evaluation using Visual 3D. Those marker sets include a modified
> Helen Hayes, the NIH model and the Oxford foot model.
>
>
>
> Alec Black
>
> Director, Shriners Gait Lab
>
> Sunny Hill Health Centre for Children
>
> Vancouver, B.C.
>
>
>
> Note from Nancy: another site reports using a Cleveland Clinic marker set
> (full or lower body only) with an AI duPont multisegment foot.
>
>
>
>
>
>
>
>
> CONFIDENTIALITY NOTICE: The information contained in this message is
> legally
> privileged and confidential information intended for the use of the
> individual or
> entity named above. If the reader of this message is not the intended
> recipient, or
> the employee or agent responsible to deliver it to the intended recipient,
> you are
> hereby notified that any release, dissemination, distribution, or copying
> of this
> communication is strictly prohibited. If you have received this
> communication in
> error, please notify the author immediately by replying to this message
> and delete
> the original message.
> Thank you
>
> ---------------------------------------------------------------
> Information about BIOMCH-L: http://www.Biomch-L.org
> Archives: http://listserv.surfnet.nl/archives/Biomch-L.html
> ---------------------------------------------------------------
>
--
Dr. Chris Kirtley MB ChB, PhD
from 1 May to 31 July 2007 I am at:
Stiftung Orthopädische Universitätsklinik Heidelberg
Leiter Ganganalyselabor
Anschrift: Schlierbacher Landstr. 200a
69118 Heidelberg
Tel: 49+06221-96 6724
Fax: 49+06221-96 6725
Clinical Gait Analysis: http://www.univie.ac.at/cga
Book:
http://www.amazon.co.uk/exec/obidos/ASIN/0443100098/203-6674734-4427132