IK joint angles unrealistic with correct model markers

[Jessica Zistatsis]

Hello,

I am modeling pediatric populations with normal and impaired gait patterns with an adapted full body model (Rajagopal 2016). The static scaled model markers match my experimental markers well for the lower extremity bony landmarks, as shown in the attached picture, but the body segments are aligned such that I'm getting high pelvic tilt, excessive hip flexion (attached picture), and maximum knee angles near 90 degrees (attached picture). I've tried adjusting the IK marker weights, but that doesn't seem to improve the joint angles. Do you have suggestions on how to improve the joint angles?

Thank you,
Jessica

[Josh Baxter]

Consider including the joint postures of the joints you know the angles to during the static trial. Including the sagittal joint angles for the hip, knee, and ankle should help improve the offsets of the gencoords during IK. This is all assuming that body scaling was pretty good.

[jimmy d]

Jessica,

Have you taken into account that the OpenSim Pelvic frame is different to non-OpenSim models? There is documentation about it here; https://simtk-confluence.stanford.edu:8 ... Kinematics

I would suggest watching our scaling youtube video which may help you scale your model;https://www.youtube.com/watch?v=ZG7wzvQC6eU&t=

[Josh Baxter]

It looks like the Rajagopal model limits the rotation range of the pelvis to 90 degrees in all directions. If your subject was standing in a direction outside of this range (with respect to your lab coordinate system) then this could also cause the problem... you can go into the osim file and change the range values for the pelvis to be 3.1459.

I've noticed this in quite a few of the osim models have this range limit on the pelvis gencoords. What's the rationale behind this?