Static Optimization Without Experimental Data

[Aaron Henry]

Hello,
I am interested in performing a static optimization on this model to determine muscle activation when a point load is applied to the head/neck. Is it possible to do this without a .mot file? I was told that I could try performing forward dynamics with all of the coordinates locked in order to determine the muscle activation required to resist the point load on the head/neck, but have not had any luck.

[Obidah Alawneh]

Hello Henry,

I am trying to do something similar to what you are doing, but my focus is on the neck/shoulder area, I was wondering if you had any luck in running the simulation without the need for a motion.

Thanks,
Obidah Alawneh

[Jacob J. Banks]

Thank you for your interest in our model.
However, if you are interested in neck modeling, I would suggest you look at a different model that has been directly evaluated for the neck:
https://simtk.org/projects/neckdynamics
Our model has been evaluated for the thoracic and lumbar spine.
To answer your question though, you 'easily' run SOpt on static tasks. You just need to create a .mot file that has no movement (and really only needs to be 1 frame long, unless you plan to vary your external forces). This is how this (and many other) model has been primarily evaluated and is really the best application for SOpt. I suggest reviewing some of the tutorials provided by OpenSim and going from there.
Thanks again for your interest. Let me/us know if you have any further questions about it (otherwise, more general inquiries may recieve more/better attention on the general OpenSim forum).
Jake Banks

[Obidah Alawneh]

Hello Jacob,

Thanks for your reply!

I am fairly new to opensim, and was wondering if you can help me with a few things.
The reason I am using your model is because I need a full-body model to obtain results on internal forces for a subject in a seated position. From what I understood from the tutorial, the best way to obtain such results is from using the Static Optimization tool.

Currently I have external forces data from another experiment, and I need to use that data as input to get the results.
But the problems I'm facing right now are:
1. I need to scale your model to match the model used in the external forces experiment, I am not sure what equipment to use to get the required data.
2. I need to create a motion that represents the seated position. But I dont know how to create a motion.

Is it possible to have a video call, to maybe help me understand what needs to be done for this simulation.

Thanks,
Obidah Alawneh

[Jacob J. Banks]

Obidah,
I'll continue to promote the OpenSim tutorials, documentation, and their forum. If you have already gone through the basic tutorials and read the documentation (as it sounds you have), I suggest >1 run through. Every time through you will learn something new (I still reference the documentation and forum).

That said, pertaining to your particular inquiries it all depends on what data you have. If you have mocap data, you can use that to scale things. If only have height and weight, can use that. If you have nothing, you should just use the base model. In terms of the 'motion' (or posture, as it sounds like it is static), if you do not have mocap data to position your model, you need to just manually place the model how you want by creating your own .mot file. One means of doing this is to adjust the model via the coordinates tab into the position you want, then run the forward sim tool to generate a motion (only for a few frames). The first line or 2 in the resulting .mot file from the forward sim tool should be the posture you desire (the rest will just be of the model free falling). Further refine things from there (as you will now have a proper .mot file to use in your SOpt). You'll then ultimitely need to apply the your external forces properly as well when using SOpt. To get the 'final' result you will input the SOpt results into the Joint Analysis Tool to calculate the joint forces of interest.

Again, I can't stress the online resources enough. Similarly, the Delp et al., 2007 paper provides a good overview of steps. There is a learning curve, but you'll get their eventually (just not overnight!). Best,
Jake Banks

[Obidah Alawneh]

Thank you Jacob, your help is really appreciated. Ill go over the documentation and tutorials again.