Hi Jalal
The best answer to your question about how the software calculates the
parameters, is to go through the tutorials and also read some of the papers
written on this issue. There are a numbers of papers listed at
http://www.anybody.aau.dk/publications.htm
I would recommend reading the paper:
Rasmussen, J., Damsgaard, M. & Voigt, M. (2001): Muscle recruitment by the
min/max criterion - a comparative numerical study. Journal of Biomechanics,
vol. 34, no. 3, pp. 409-415.
Your description on the how the process work is absolutely correct.
Here are some answers to the numbered items:
1, yes your are right the fullbody model is available and can be used for
this kind of model
2, the best thing will be to drive the model using motion capture data
having a recording with and without the harness. Even if you had the motion
driven 100% correct and you had 100% realistic boundary conditions between
the feet and the ground, there would still be a balancing issue. The
anthropometry of the model will never be exactly as the subject so if you
impose he same motion but a slightly different mass distribution among the
segments in the model, it will impose forces on the model which are not
there in real life. These forces can be so large that the model “tip over”
so if the model is walking and only one foot is on the ground these errors
could be so large that there are no way the contact between the foot and the
ground can balance it unless the foot is “glued” to the ground. Introducing
the weak muscles is basically introducing this glue, but it will only make
use of it if there are no alternative.
-
This is not the most complicated part, in principle you will need to have
at least a three point connection. So you will need one point on the heel
and a point medial and lateral on the front of the foot. This will be like
a tripod contact, it is a simplification, you may need to introduce more
points, I have never tried this for walking.
-
If you are sure the force is constant, it is easy to apply to the pelvis
I guess.
Best regards
Søren, AnyBody Support
From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of j.arabneydi@ece.ut.ac.ir
Sent: 20 February 2008 11:03
To: anyscript@yahoogroups.com
Subject: [AnyScript] gait maodel ability questions
Hi,Søren
speacial thanks for your attention and attempt to answer my questions.
how does software calculate parameters?
I think,at first a full body model are developed,after that data are
aquired from a human gait such as (position,velocity,…via markers) and
they are applied to model.Now,another parameters can be calculated such as
internal forces,by knowing structure and real data,doesn’t it?
you mentioned 4 problems that are disability reasons to calsulate forces:
1)a full body model is needed.I think a full body model exists in any
body,doesn’t it?
2)balancing issue,I can not understand this problem because I dont know
where this model comes from!Is it a human model which is walked by real
data or is it a passive model!if it is a passive model,how can it
walked?by which data?and if it uses human data,why we should need a
“helping hand”?
3)If we have a full body model like that is existed in any body,does we
have to do anything more?if yes,is it very complicated?
4)If we now harness force is constant and known,is there any thing else to
do?
> Hi Jalal
>
>
>
> I will try to answer your questions:
>
>
>
> 1) In the Gait3D model from the repository the force has been
> measured
> and is applied to the model.
>
>
>
> In principle it is possible to do a gait model where the reaction forces
> are
> calculated, but it is a rather challenging model there are a number
> problems:
>
>
>
> 1) You will need the fullbody model, if using only the lowerbody the
> calculated reactions will reflect this.
>
> 2) There is a balancing issue, you will need something to help the
> model keep the balance and this “helping hand” may affect the calculated
> reactions.
>
> 3) You will need to setup shifting boundary conditions between the
> feet
> and the ground, for details on how to do this please have a look in the
> PedalDemoConditional model, in the JointAndDriver.any file there is a
> example of such conditional force.
>
> 4) When the model is running with the harness you will need to know
> something about the forces it apply to the pelvis.
>
>
>
> So as you can see there are some issues I am mostly concerned about item
> 2.
> My first suggestion on how to solve this is to add a set of weak muscles
> between pelvis and the ground. If the model is having difficulties to hold
> the balance then these muscles will become active. If there are no
> problems
> they will be inactive because they are weak and will lead to high activity
> if activated.
>
>
>
> There might also be an issue about the motion and how much it will be
> affect
> by the harshness.
>
>
>
> The model would be better if you could measure the ground reaction forces.
>
>
>
> If you have not done so I will recommend you going through the tutorials
> first, it will make it easier to do this model, but it will still be a
> difficult one.
>
>
>
> Best regards
>
> Søren, AnyBody Support
>
>
>
>
>
> _____
>
> From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
[mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com] On
> Behalf
> Of j.arabneydi@ <mailto:j.arabneydi%40ece.ut.ac.ir> ece.ut.ac.ir
> Sent: 19 February 2008 11:26
> To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> Subject: [AnyScript] gait maodel ability questions
>
>
>
>
> Hi,
> I am Jalal.Arabneydi,MSc student of electrical engineernig(control) at
> University of Teharn.My thesis is about gait therapy via an inetelligent
> control system that has an ability to balance and assist the patient to
> walk on treadmill while patient is protected and ballanced via a BWS(Body
> Weight Support) system.
> I have some questions about peresented gait model in anybodytech.
> 1)I think,this software via markers which mounted on body,calculates the
> shape and by having the mass of segments and muscles and position data,the
> reaction ground forces are achived,doesn’t it?or are there invidual
> sensors for sensing force?
> MY PLANT:patient is waliking on treadmill while it is been unloaded via a
> harness.
> 2)to investigate the affect of unloading system in reaction ground
> forces,can we change some parameters or settings to achive the reaction
> groung forces in differnet unloading?(known unweighting).in other
> words,can we apply a force to pelvis against gravity direction so that
> software can caluculate new forces with previous data set or previous
> scenario?
> I’m looking forward to your response.
>
> Jalal.Arabneydi
> Best Reagards.
>
>
>
>
>
> [Non-text portions of this message have been removed]
>
>
[Non-text portions of this message have been removed]