the role of local marker optimization

Dear Colleague in Anybody,

Hi, I try to follow the discussion on this forum, and I could not find any of them who have discussed about optimization function of local marker in Anybody.

I have a question regarding optimization process, what is actually the effect of optimization process for our local marker when we do the initial guess?
I was wondering if the effect of optimization is for defining the trajectory of dedicated marker or maybe I thought wrong. Is there anybody from anybody can share this knowledge?
Below is the example of optimization code in modelsetup.any file

CreateMarkerTD LTHI (
MarkerName=PrefixDef(LTHI) ,
MarkerPlacement=Left.Leg.Seg.Thigh,
OptX=“Off”,OptY=“On”,OptZ=“Off”,
WeightX=1.0,WeightY=1.0,WeightZ=1.0,

Model1=MotionAndParameterOptimizationModel, Model2= InverseDynamicModel
) = {
sRelOpt= {0.01801, -0.230012, -0.0700};
};

======================

In that example, what is the different effect that might happen if we optimize in x, or y, or z direction?

I am using AMMRV.1.3.1 (anybody version 5.0)

Thanks for sharing

Adhi
Groningen University, netherlands

Hi,

Let’s see the effect of ‘OptX=“Off”,OptY=“On”,OptZ=“Off”’ in CreateMarkerTD class.

By default, sRelOpt values are {0.01801, -0.230012, -0.0700}.

Now you set only OptY=“On”, so the Y component of sRelOpt can be changed by optimization. Thus X and Z components of sRelOpt would not be changed.

I hope this may help to you.

Best regards,
Moonki

Dear Moonki,

Thanks for the reply, but can you give me more explanation on : what is the effect of markers being optimized? as sometimes, without optimization, the model still can run without any error message. is there any different in the result at the end?? for example, on muscle activity or muscle forces value between optimized markers and not being optimized ones??

Regards
Adhi
Groningen, The Netherlands

Hi,

Let me explain further about this subject.

When you use GaitFullBody example in the repository,
you can find ‘Main.Studies.ParameterIdentification’ in the model.

And there are several parameters to be optimized.
->PelvisWidth, ThighLengths, ShankLengths, FootLengths, TrunkHeigh, UpperArmLength, LowerArmLength and the components of Marker locations.

So if you would not change the position of markers, then the optimization solver would try to change the length values of the segments.

But there are several markers that we may not guess their initial positions well such as LTHI and RTHI. So for this kind of markers, I would recommend you to allow their positions to be changed.

So my answers is that according to your settings, the optimized lengths of human segments will be changed. And of course, the resultant kinematics will be changed.

Best regards,
Moonki

Thanks Moonki for the explanation, that really helps me.

Regards
Adhi