AMMR4 beta Kinematic analysis failed

Main Forum
8

Dear Zyax :slight_smile:

  1. I think you should use your joint drivers with very low weights in the marker tracking and with higher weights in the parameter identification. So in the marker tracking study, your markers have much bigger weights and will change the spinal angles.
    Regarding the OptX, OptY, OptZ of the markers, I have explained it in another topic called " Segments length optimization", which you can read about it :slight_smile:
    Remember that the rhythms and the joint drivers are constraining the joint angles and are the same basically. So if you want to use the rhythms, you should also have very low weights for them. But the joint drivers are better to make a better guess of the spine if you have the spine alignment measured from somewhere like X-rays. If you are modelling a normal spine and you do not have extra spine alignment data, you should drive the joint drivers to zero, which is exactly what the rhythms will do when you drive the rhythms to zero :slight_smile:

  2. It seems that the hand marker position on the model and your data does not match.
    If the hand is not very important to you, you can exclude the the HAND_LENGTH optimization in the LabSpecificData.any.
    If you need it, then you probably need to change the marker position in the model to tip of the fingers or something that fits the anatomical explanation of where you out the marker in the lab.

  3. In your picture, I can see that the red and blue markers in lumbar are not close to each other. So the markers are showing almost a straight guy and the spine alignment is a curved one. And I can see that the OptZ for lumbar markers are OFF. So the markers cannot move in that direction (which I think that is correct to have it this way) and the spine cannot move because you are driving it to special spine angles (which I also think that is a correct way to do it). However you should seek the problem on the marker positions in the lab data and also the spine alignment. I can see that they do not match, and that is probably why it gives results with big errors.

  4. Maybe it is better to solve the above issues first, and then run 'RunAnalysis'. But when you did that, let me know if you still see the kinematic error and you see it in which study :slight_smile:

Best regards,
Hamed