Hello everyone,
I am working on a biomechanical modeling project using OpenSim software to estimate lower limb joint angles. For this, I have placed markers on the lower limb based on plug-in-gait lower limb protocol. However, during my experiments, I placed markers on a stick that was 2-4 cm away from the skin instead of attaching them directly to the skin.
My question is: should I give weight zero to these markers during scaling, or should I manually adjust their virtual position to match the experimental markers on the tibia and thigh?
Given the fact that these markers were placed on a stick and not directly on the skin during the experiments, what would be the best approach? How much weight should I give to these markers during scaling and during IK analysis? And should I manually adjust their position during scaling?
I would really appreciate any suggestions or insights on this matter. Thank you in advance.
markers weight and adjusting virtual markers position
- Mohammadreza Rezaie
- Posts: 407
- Joined: Fri Nov 24, 2017 12:48 am
Re: markers weight and adjusting virtual markers position
Hi, I suggest disabling the tracking markers entirely.
https://simtk-confluence.stanford.edu:8 ... leshootingRely on markers that correspond to anatomical landmarks and functional joint centers (FJC) to position and scale the generic model.
- John Davis
- Posts: 59
- Joined: Mon Aug 26, 2019 7:42 am
Re: markers weight and adjusting virtual markers position
I agree with Mohammadreza, you don't want to use those markers at all during scaling. They should be ok to use during IK, though. There's no need to to manually adjust the marker position--just make sure the <fixed> flag is set to false for that marker in your .osim model file, and when you do scaling (with the "Adjust model markers" option checked), the marker registration step of the scaling tool will automatically move the markers on the model to the correct location, using your experimental marker data after posing the model.
Re: markers weight and adjusting virtual markers position
Thanks a lot for your help... in which cases should I adjust the marker positions manually during scaling?
https://simtk-confluence.stanford.edu:8 ... leshootingAgain, avoid adjusting the positions of the landmark and FJC virtual markers to match the experimental markers.
- John Davis
- Posts: 59
- Joined: Mon Aug 26, 2019 7:42 am
Re: markers weight and adjusting virtual markers position
You should not typically have to manually adjust (in the literal sense of clicking and dragging in the GUI) any markers at all, except for when you place the virtual markers on your generic model for the very first time.
The scale tool is actually two different tools in one: first, it scales the segments of the generic model proportional to the marker pairs you specify in the scale tool setup config file (or manual scaling factors instead--marker data are not strictly necessary for this step, but in practice marker data are almost always used to do this step of scaling).
Second, and completely separate from the first scaling step, the scale tool performs one frame of inverse kinematics on the scaled model, then "registers" (moves) every virtual marker with the <fixed> attribute set to "false", so that the virtual markers match the experimental marker locations. These non-fixed markers are usually your tracking-only markers, like a thigh cluster.
Markers with the <fixed> attribute set to "true" are not moved during this second step. Typically, these fixed markers are at anatomical landmarks like the ASIS and or the medial and lateral malleoli. These fixed markers are also usually the ones you use for the first step of scaling as your marker pairs, and the ones you use for doing the single frame of IK that happens at the beginning of the second step of scaling.
This OpenSim webinar cleared up a lot of my own confusion about the scale tool: https://www.youtube.com/watch?v=ZG7wzvQC6eU
The scale tool is actually two different tools in one: first, it scales the segments of the generic model proportional to the marker pairs you specify in the scale tool setup config file (or manual scaling factors instead--marker data are not strictly necessary for this step, but in practice marker data are almost always used to do this step of scaling).
Second, and completely separate from the first scaling step, the scale tool performs one frame of inverse kinematics on the scaled model, then "registers" (moves) every virtual marker with the <fixed> attribute set to "false", so that the virtual markers match the experimental marker locations. These non-fixed markers are usually your tracking-only markers, like a thigh cluster.
Markers with the <fixed> attribute set to "true" are not moved during this second step. Typically, these fixed markers are at anatomical landmarks like the ASIS and or the medial and lateral malleoli. These fixed markers are also usually the ones you use for the first step of scaling as your marker pairs, and the ones you use for doing the single frame of IK that happens at the beginning of the second step of scaling.
This OpenSim webinar cleared up a lot of my own confusion about the scale tool: https://www.youtube.com/watch?v=ZG7wzvQC6eU