Wrist markers and its awkward rotation

Happy new year!

Your assist for solving the wrist markers’ problem was very helpful.
In the similar way, I was trying to modify the left wrist markers.

In this case, I did not have reference papers for each marker’s location, thus I set the position as a neutral posture and found the relative marker position manually.

  
CreateMarker  LWRA (
  MarkerName=LWRA,  
  MarkerPlacement=Left.ShoulderArm.Seg.Radius,
  OptX="Off",OptY="Off",OptZ="Off",  
  WeightX=1.0,WeightY=1.0,WeightZ=1.0,

  Model1=MotionAndParameterOptimizationModel,Model2= InverseDynamicModel
  ) = {
    sRelOpt = {-0.18-0.07-0.00, 0.07-0.01+0.003, 0.01};
  };
  
  CreateMarker  LWRB (
  MarkerName=LWRB,  
  MarkerPlacement=Left.ShoulderArm.Seg.Ulna, 
  OptX="Off",OptY="Off",OptZ="Off",  
  WeightX=1.0,WeightY=1.0,WeightZ=1.0,

  Model1=MotionAndParameterOptimizationModel,Model2= InverseDynamicModel
  ) = {
    sRelOpt = {0.03-0.03-0.003, -0.205-0.07, 0.01};
  };

This works well for some subjects and some motions, but sometimes it has such malfunction as the attached picture.
When the motion and parameter optimization begins, the wrist flips upside down while the wrist stays still. So, as you can see in the picture, the thumb goes to the ulna side, and the palms goes to upper.

I think the hand segment follows the markers awkwardly, how can I handle it?
Thanks in advance.

Sincerely yours,
Sheen Dong-Pyoung

Hi,

Would you upload your model? Then we may be able to look it in detail.

Best regards,
Moonki

Here is the analysis model.
Thanks. :slight_smile:

Hi,

Let me explain what I modified on your model.

  1. I modified the definitions of LWRA and LWRB markers as follows:
  CreateMarker  LWRA (
  MarkerName=LWRA,  
  MarkerPlacement=Left.ShoulderArm.Seg.Radius,
  OptX="Off",OptY="Off",OptZ="Off",  
  WeightX=1.0,WeightY=1.0,WeightZ=1.0,

  Model1=MotionAndParameterOptimizationModel,Model2= InverseDynamicModel
  ) = {
     [b]sRelOpt = {-0.18-0.05, -(-0.07+0.02), 0.01};[/b]
  };
  
  CreateMarker  LWRB (
  MarkerName=LWRB,  
  MarkerPlacement=Left.ShoulderArm.Seg.Ulna, 
  OptX="Off",OptY="Off",OptZ="Off",  
  WeightX=1.0,WeightY=1.0,WeightZ=1.0,

  Model1=MotionAndParameterOptimizationModel,Model2= InverseDynamicModel
  ) = {
    [b]sRelOpt = {0.03+0.01-0.005, -(0.205+0.06), 0.01};[/b]
       }
  1. Also, it seems that the initial values of WristFlexion and WristAbduction are important to get the right orientation of the hand segment.
    So I modified those values in the ‘TrialSpecificData.any’ file as follows:
    AnyFolder Left ={
      
      AnyVar GlenohumeralFlexion =15; 
      AnyVar GlenohumeralAbduction = 10; 
      AnyVar GlenohumeralExternalRotation = 00; 
      
      AnyVar ElbowFlexion = 95; 
      AnyVar ElbowPronation = 80.0; 
      
     [b] AnyVar WristFlexion = 60;
      AnyVar WristAbduction = 60;[/b]

      AnyVar HipFlexion = -20.1819;
      AnyVar HipAbduction = -3.050171;
      AnyVar HipExternalRotation = 0*-16.907892;
      AnyVar KneeFlexion = 7.764009;
      AnyVar AnklePlantarFlexion = 48.790382-30;
      AnyVar SubTalarEversion =0.20658;
    };

And I attached the modified model.

I hope this may help you.

Best regards,
Moonki

Thank you for your assistance.

However, I am somewhat curious about the initial position and the wrist’s kinematics because wrist kinematics was not severely depending on the its initial position in last analysis trials.
So I’m changing the wrist joint angle and looking how its kinematics is determined.

In my thought, the wrist could be positioned properly according to the wrist and hand markers when the wrist markers were defined on the hand segment. However, by changing its definition on the radius and ulna, the hand segment’s position is determined only by the hand marker. So I believe this is the reason of the awkward hand position.
Am I right?

Hi,

Yes. Your guess is right.

  1. If you define the LWRA and LWRB markers on the hand segment, the kinematics of the hand segment can be solved without any problems.
    But then, the kinematics of unla and radius segments may be problem.

  2. If you define the LWRA and LWRB markers on the radius and ulna segments respectively, then the kinematics of ulna and radius segments can be solved without big problems.
    But then the kinematics of the hand segment can be solved mainly with the LFIN marker so there may be more than one answer by AnyBody’s kinematic solver.
    This is a kind of bifurcation that general kinematic problem may contain.
    It means that there are multiple solutions that satisfy the kinematic constraints in the model.

Whatever marker definition that you may use, the initial position of the human segments are important to get the right answer that you may want to see.

So there may be no royal way to do this. Just select the proper marker set that may be most stable in your analysis, and find a good initial guess of the joint angles.

I hope this may help you.

Best regards,
Moonki