Markers on leg6dof9musc

[Misha Patel]

Hello,
I am trying to collect experimental data as explained in this section (http://simtk-confluence.stanford.edu:80 ... ental+Data). I am focusing on the knee (femur and tibia) and I am trying to distinguish where to put the markers. When I look at the leg6dof9musc.osim model and focus on the following markers:
R.Thigh.Upper
R.Thigh.Front
R.Thigh.Rear
R.Shank.Upper
R.Shank.Front
R.Shank.Rear
Have these markers been placed arbitrarily on the thigh and shank? How could I replicate these markers on an experimental body?
Also, if i wanted to amend the model dimensions based on data I have collected, should I follow these markers or could I use different marker placements?
Thank you

[Bryce Killen]

Hi Misha,

There a couple of questions here and hopefully I have interpreted them correctly although they are related

Have these markers been placed arbitrarily on the thigh and shank? How could I replicate these markers on an experimental body? Also, if i wanted to amend the model dimensions based on data I have collected, should I follow these markers or could I use different marker placements?

You do not necessarily need to replicate these model markers, as you suggested you can create your own marker set on the generic leg6dof9musc opensim model to suit where you will be placing your experimental markers. This may also resolved the first question about how to replicate them in experimental data collection.

To sum up, if you are not confident in replicating the model marker placement in an experimental data collection you can create your own custom marker set on this model to match landmarks on your experimental participant. By placing these markers on a generic model you can more easily replicate them in a experimental setting because you know where the markers were intended to be placed.

Hope this helps,

Bryce

[Misha Patel]

That's great, thank you Bryce.
The data I am trying to collect is in order to use the scaling tool. From my understanding I will need to create a '.trc' file. Do I need to motion capture data or are photographs of static poses enough in order to find a scale length. I ask this because on the 'Getting started with Scaling page' it says the following with regards to data collection

"Data Collection and Other Preparation:
When collecting data, take pictures of your subjects in the static pose. These pictures are valuable for evaluating the results of the Scale Tool."

But when I look further into 'Marker (.trc) Files' in the User's guide it talks about sample rates and motion capture trials.

If I am to use one of the tools listed on this page (http://simtk-confluence.stanford.edu:80 ... otion+Data) what experimental data am I required to obtain?

Thank you,
Misha

[Bryce Killen]

HI Misha,

From what I understood you are just wanting to scale a model you already have. To use the scale tool there is two options . It might be worth having a look here if you haven't already http://simtk-confluence.stanford.edu:80 ... th+Scaling


1) Manual scale factors , if you have actual physical measurements of your subject ie. femur length, width etc. You can calculate manual scale factors based on the known value (your subject) and the model value.

2) The more common is to use a motion capture trial. You can use a motion capture trial to create a .trc file which is a trace file which contains the coordinates of each of your markers. These trc files can be created either using btk (https://simtk.org/frs/?group_id=660) or MOtoNMS (https://simtk.org/projects/motonms/). You then define a number of marker pairs which define what is called a measurement. This measurement is calculated for both the known (experimental) markers and model markers . This then creates a "scaling factor" based on the ratio of known value to model value. This value is then used to scale a specific body (femur, tibia) in a given dimension (x,y,z) depending on how you set up scaling to work.

Hope this clears it up

Cheers
Bryce

[Misha Patel]

Hello,
Yes, that is exactly what I am trying to do. Thank you, I had already seen the link you posted and have used that as a starting point. Thank you for clearing up that there are 2 separate methods.

Just to clarify:

Following the first method, I would simply need to take measurements of the femur etc and then I would calculate individual scale factors for different marker pairs? Would I need to create a trc file for this method? If not, how would I implement the scale factors on my model in Open Sim... would I still be using the Scale tool?

For the second method, Do I need to place my markers on a subject and take a video of them walking? I have already defined my marker pairs and can take measure the distance between them to define a scale factor for my subject in a static pose. I am struggling to understand how the motion capture can be analysed and manipulated into a trc file.

It seems as thought both methods would simply require static photos in order to measure various lengths which leads to confusion as to why and how you would use a motion capture trial in the second method?


Thank you,
Misha

[Bryce Killen]

HI,

Following the first method, I would simply need to take measurements of the femur etc and then I would calculate individual scale factors for different marker pairs? Would I need to create a trc file for this method? If not, how would I implement the scale factors on my model in Open Sim... would I still be using the Scale tool?

So you would need to know the actual length of the femur lets say it is 45cm so 0.45m , then if you know the measurement of model lets say it is 40 cm so 0.4m . Your scale factor for the femur Y would be 0.45/0.4 = 1.125 this can be input in the scale tool as a "MANUAL SCALE"

For the second method, Do I need to place my markers on a subject and take a video of them walking? I have already defined my marker pairs and can take measure the distance between them to define a scale factor for my subject in a static pose. I am struggling to understand how the motion capture can be analysed and manipulated into a trc file.

So you need to capture them using a motion capture system ie Vicon which will save the markers as a c3d file etc. then you can use the tools I sent before btk/MOtoNMS to create your trc file. Using this in the scale tool where you already have your marker pairs defined will automatically calculate the scale factors. SO you need motion capture trial not jsut a video/static photo because you can't measure lengths confidentially.

Cheers

[Misha Patel]

Hi,

Thank you so much for clearing that up. I hadn't previously seen the manual scale section of the scaling tool.

If I am to use manual scaling, how can I obtain the measurements of an existing model in Open Sim? Is there a data set for each OpenSim Core Model? I am working on the leg6dof9musc.osim file that came with OpenSim when I dowloaded the software

Thank you

[Thomas Uchida]

how can I obtain the measurements of an existing model in Open Sim?

There should be an example ScaleTool setup file for the 2354 model in the OpenSim Models directory (e.g., C:\OpenSim 3.3\Models\Gait2354_Simbody). You can find additional information here:
- "Scaling" pages in the Confluence documentation: http://simtk-confluence.stanford.edu:80 ... im/Scaling
- "Tutorial 3 - Scaling, Inverse Kinematics, and Inverse Dynamics" on Confluence: http://simtk-confluence.stanford.edu:80 ... e+Dynamics
- "OpenSim Webinar: Tips and Tricks for Data Collection, Scaling and Inverse Kinematics in OpenSim" on YouTube: https://www.youtube.com/watch?v=ZG7wzvQC6eU

[Misha Patel]

Hello,
I've managed to obtain some data from a motion capture trial. It is displayed in an excel file.
Can I convert this into a trc file in order to scale a model?

[Thomas Uchida]

Can I convert this into a trc file in order to scale a model?

Please see the "Marker (.trc) Files" page in the Confluence documentation (http://simtk-confluence.stanford.edu:80 ... c%29+Files).