Gait motion with c3d file

Hi, I have tried to analize gait motion by using c3d file.
I’ve asked many questions and solved some problem.
But I don’t know what is the problem of this analysis.
So I hope that somebody help me to solve this problem.

First file is AMMRV1.1 version and some muscle’s activity is too high.
I cannot find what the problem is.

Second file is AMMRV1.2 version and there is kinematic problem.
Could you tell me what the problem is?

Please help me.
Thanks.

  • Killian -

Hi Killian,
I had a brief look at your models. In model one there seems to be a problem with the wrapping of the Iliacus muscles on the pelvis. Please investigate the problem. You can find the definition of the wrapping surface in the muscle files. Visualize the wrapping cylinders and see if the muscles wrap nicely, if not you may have to change the inital vectors of the wrapping muscles or the geometrical definition of the surface a bit, if the wrapping surface is e.g. to short some muscle may just “fall of the edge”. A similar problem is in Model two in the shoulder. These kind of problems can appear with scaling (I think your models are quite “skinny”) and unflavourable shoulder positions. In model one another problem seems to be in the ankle. Switch on the reactions in the Subtalar joint. If you are not interested in the ankle region this may help and be ok.
I could not see a kinematic problem in model 2. It runs ok here. I hope this helps a little.
Best regards,
Sebastian

Thank you very much.
I will try to do what you said.

And the second one works well as you said.

I have one more model and this one has a kinematic error.
Could you look at this file and tell me what the problem is?

Thanks.

  • Killian -

Hi,

i have launched the optimization but i could not see any kinematic error. Can you be more specific on what the problem is?

However what i could see is that you have too many design variables. Try to optimize less parameters. I think some more marker locations should be fixed.

Best regards, Sylvain.

Thanks.
I solve the problem as using C3D editor.
But sometimes I have an error someting like below.

  1. InitialConditions (Operation: Main.Studies.KinematicStudyForParameterIdentification.InitialConditions):
  2. …Design variables have been updated.
  3. …Load time positions have been re-established.
    Failed to resolve kinematic constraints. Newton relaxation too small. (final kin. error = 2.634118E-003)
    Constraint no. 14 above error tolerance 0.001000, error = 0.002634.
    Constraint no. 15 above error tolerance 0.001000, error = 0.001817.
    Constraint no. 161 above error tolerance 0.001000, error = 0.001480.
    ERROR(OBJ.MCH.KIN3) : D:/A…y/A…)/A…2/A…2/A…n/E…s/G…n/GaitFullBody.main.any : KinematicStudyForParameterIdentification.InitialConditions : Kinematic analysis failed in time step 0

I had lots of this problem before as well.
In this case, it happened at initial conditions step but it happens at the kinematics step too.
I hope to know what the problem is and how to solve the problem.
I attatched the file so check the file please.
Thanks.

  • Killian -

Hi Killian,
it is mostly related to a bad initial guess on the postion, so bring the model closer to the markers may help (done in the Trailspecificdata file). Also changing the inital segment length will help.
Best regards,
Sebastian

Thanks.
But I can see a similar error message during Kinematic analysis after overcoming this initial problem.
The message is something like following.

  1. Kinematic analysis…
    Failed to resolve kinematic constraints. Newton relaxation too small. (final kin. error = 1.009475E-003)
    Constraint no. 14 above error tolerance 0.001000, error = 0.001009.
  2. …Kinematic analysis terminated
    ERROR(OBJ.MCH.KIN3) : D:/A…y/A…)/A…2/A…2/A…n/E…s/k…t/GaitFullBody.main.any : KinematicStudyForParameterIdentification.Kinematics : Kinematic analysis failed in time step 9

In this case, how can I solve this problem?
Do you think it has a problem at the initial condition as well?

The intial conditions will not affect what comes later, but the segment length may. So try a better guess on the initial segment length.
Best regards,
Sebastian