Will changing the joint of a scaled model, affect the scaling

[Radhakrishnan Vignesh]

Hello,

Hope you are doing well. I have a question. I am using a dataset where I have a scaled model and I would like to know if I change the joints of the scaled model (editing the osim file), would it invalidate the scaling?

Thanks

[Radhakrishnan Vignesh]

Hi,

Any luck with the above question? I tried to change the knee joint type of a scaled walker model (gait 2392) from coupled to ball and got the dislocation of the knee joint centre (image attached).

It is due to the coupling fuctions for the y-direction translation. How do I change it?

Capture.JPG (78.44 KiB) Viewed 711 times

[Thomas Uchida]

The highlighted text defines how the joint's child frame translates in the y-direction relative to the joint's parent frame as the knee_angle_l coordinate changes. If you wish to remove this function, you could (1) pick a value for knee_angle_l at which the new model will have the same joint configuration as the old one, (2) evaluate the SimmSpline at this "x" value, and (3) set the y-axis offset of the child frame to the SimmSpline's "y" value. The code that evaluates a SimmSpline can be found in the SimmSpline::calcValue() method here: https://github.com/opensim-org/opensim- ... e.cpp#L511.

[Radhakrishnan Vignesh]

Thank you. I have one more question:

Does changing the joint type of a scaled model invalidate the scaling of the model ?

Thanks

[Thomas Uchida]

Does changing the joint type of a scaled model invalidate the scaling of the model ?

It depends how you are validating your scaled model. Modifying a joint will change the model's kinematics, so presumably any validation based on the model's kinematics would need to be repeated.

[Radhakrishnan Vignesh]

Thnak you.

The reason I ask is that when scaling a model with free joints, the bones look dislocated despite very small scaling error. I am not sure how to determine if this is correct.

Additionally, is there a function in OpenSim which detects dislocations or abnormal gait (not pathological)? For example, when changing the marker locations or joint constraints, the bones appear to be moving abnormally (especially for the free joint). Since I have 500 subject data, it is not efficient to visually verify them all.

Thanks