Page 1 of 1

Subject-specific geometry

Posted: Thu Oct 29, 2020 1:01 am
by akoussou
Dear OpenSim experts,

I'm using Rajagopal model to simulate gait in children with cerebral palsy. In this population orthopaedic deformations can be present, for example, an excessive femoral anteversion/tibial torsion. These two properties are measured and known.

Thus, I would like to know if there is a way to adapt the lower limbs of my model to take into account of these deformities. And if this adaptation will have an impact on the differents muscle-tendon lengths computed because I have already read on the forum that meshes are only for "visualization". What does this exactly mean?

To achieve my goal, I have already, in a first time, tried to adapt the geometry file of the femur :

- I have modified l_femur.vtp. I have changed the third column of <PolyData> - <Piece> - <PointData> - <DataArray>. I have called this new file l_femur2.vtp.
- I have placed this new file in my Geometry folder.
- Then, I have edited my .osim file and replaced my l_femur.vtp by l_femur2.vtp.

But, nothing changed. My new model looks exactly the same that the older one.
How is it possible ?

Sorry for all these questions, I would be so grateful if you could take the time to answer all.

Thank you in advance,

Regards

Re: Subject-specific geometry

Posted: Thu Oct 29, 2020 2:30 am
by akoussou
Update :

I finally suceeded to modify the geometry of my femur. I deplaced it along the Z-axis and it appeared to me that it did not change anything about the muscle-tendon units, which stay attached to the same points. (cf. Figure).
So I better understand what I could read on the forum when OpenSim team say that geometry is "only for visualization".

Thus, my goal is now, to understand if there is a way to take account of the subject bones morphology. Especially for the calculation of muscle-tendon units lengths.

Thanks in advance,

Regards.

Re: Subject-specific geometry

Posted: Fri Oct 30, 2020 1:52 pm
by ongcf
With regards to muscle-tendon lengths, you could look to redefine any properties within the GeometryPath property of each muscle. Perhaps one way to iterate is to make changes to the visualization of the bones, and then change PathPoint's within the GeometryPath so that the muscle-tendon lines appear to be touching the perimeter of the bones.