I’m trying to figure out the CreateMarker class because I have to create a new marker set. The value “sRelOpt” is “coordinates of the marker unscaled” according to the example files in the repository (AMMV1.1 & Gait…). Now, my question is: what does that mean?
When checking the scaling laws used (in the example files) I find the “old joint-to-joint” method (etc) and AnyMan.any. In AnyMan.any the overall height is set to 173 cm (and weight is 63 kg). Does that mean that the marker set (concerning the sRelOpt) is based on a subject of 173 cm?
Does sRelOpt scale depending on subject height (etc) for the PiG marker set? And how?
When creating a new marker set that uses a few of the marker positions from the PiG, is the old sRelOpt relevant (when the subject is 10 cm higher and 25 kg heavier)?
The geometrical scaling of the GaitLowerExtremity model is actually done by the optimization as the segments length are optomized. The body height in the AnyMan file of this model is only used to calculate the BMI.
So the sRelOpt is the location of the marker on the unscaled body model, corresponding to the ScalingStandard law. This sRelOpt is then scaled together with the bone. So sRelOpt is valid for all subject size for one marker set.
...
So the sRelOpt is the location of the marker on the unscaled body model, corresponding to the ScalingStandard law. This sRelOpt is then scaled together with the bone. So sRelOpt is valid for all subject size for one marker set.
Best regards, Sylvain.
I know the marker positions in relation to a couple of bony landmarks on my subjects. So, what I do (when implementing my "new" marker set) is:
[ol]
[li]Scale a body model according to one of my subjects.[/li][li]Define the new marker set in ModelSetup.any with sRelOpt values corresponding to marker position in relation to the segment (that is scaled to my subject).[/li][li]Run Parameter Identification etc...and I should get the motion data needed for inverse dynamics analysis (if the marker set was "good enough" for the body model used).[/li][/ol]
For my next subject (or next motion with the same subject) I should be able to skip step 1 and 2 (except setting the correct height and weight) and go directly to step 3.
Do I understand the "workflow needed" for implementing a new marker set?
Because I don't know where the markers are in relation to any unscaled body model, I don't see how I can do without step 1 when implementing a new marker set. And, if scaling the body model I don't see that I can "reuse" the sRelOpt values from your PiG marker set even though I use some of the markers from PiG.
I thing you understoud well the work flow for introducing a new marker set. I have just one coment on the steps 1 and 2, it may not be strictly necesary to scale the model according to your subject before placing the new markers.
If for example you know that the marker is directly on a bony landmark, let’s say on the malleolus, then you can place it visualy on the model without prior scaling. Or use some of the already defined bony landmarks of the model.
If the marker is placed for example “10 cm above malleolus” for your specific subject size then it is ok to presclae the model to use the correct distance.
Howerer remenber that most of the markers that are on the midle of a bone have their location optimised in the 3 directions, and the ones on clear bony landmark are fixed. So in most cases you can just place the fixed ones using the predifined bony landmark and place roughly the ones that will be optimised without prior scaling.