OpenSim vs Vicon joint kinematics and kinetics

[Philippe Dixon]

Dear all,

I have collected motion capture data using the Vicon Plug-in Gait model and have used that marker set to create simulations in OpenSim. In order to validate the results, or at least gain more confidence in the simulations, I have been comparing both the joint kinematic and kinetic outputs across platforms. The kinetics match very well, but the pelvis and hip kinematics produce a consistent offset of about 15 degrees in the sagittal plane across systems. To note, the measured muscle activity (EMG) also match well with the simulation results. Has anyone come across this problem or have any suggestions as to why this disagreement occurs?

[Yusuke Okita]

Hello,

That offset might be caused by the different definition of the pelvis anterior tilt between .osim model (e.g., gait2392) and PiG model. This page might help you http://simtk-confluence.stanford.edu:80 ... negeometry

Pelvic Tilt and Hip Flexion Angle (Gait2354 vs. Clinical Measurements)

The pelvic frame is located midway between the two ASIS. In the neutral position, the model has zero pelvic tilt with respect to ground (or lab). In other words, the pelvic frame and ground frame are aligned at neutral (pelvic tilt = 0 degrees). In many clinical papers, neutral corresponds to 12-13 degrees of pelvic tilt. This will lead to an offset when comparing pelvic tilt and hip flexion angles from the gait2354 model to some data from clinical papers or gait analysis lab data.

To compare angles from the gait2354 model to clinical data, you can subtract the angle formed between the horizontal plane and a line between the ASIS and PSIS markers from the clinical pelvic tilt measurements. You should also add this value to the clinical hip flexion measurement. If angle formed between the horizontal plane and a line between the ASIS and PSIS markers is not known, a value of 12-13 degrees is typical.

Best regards,
Yusuke

[Philippe Dixon]

Thanks Yusuke, that is exactly what I was looking for. Match is much improved.

Phil

[Mireia Claramunt]

Hello,

I am performing gait analysis (kinematics and dynamics) with gait2392 model from OpenSim. The data was captured through a set of infrared cameras, using the Plug-in-gait marker model. Computing the kinematics, I have the offset problem that Phillippe commented previously, which I solved applying the 12-13 degrees solution given. I understand that you substract the same degrees in pelvis tilt than in hip flexion.

My question appears when I do inverse dynamics. Should I load the .mot file created in the first instance by OpenSim, the one that contains the offset issue, or should I make a new .mot file with the rearranged pelvis tilt and hip flexion angles and then compute the inverse dynamics?

Best regards,

Mireia

[jimmy d]

You should use your original model. Applying the kinematic offset is only for reporting purposes.

[saylee patkar]

Hello,
I am performing gait analysis on gait2354 model with data collected from vicon nexus system. I am also having some problem with comparison of kinematic data.
The problem is with knee joint and I am facing the problem in only one knee joint. The offset is almost 20-25 degrees.
What can be the reason for this?