Muscle strength scaling using measured joint torques

Main Forum
1

Hi everybody,

I am trying to scale the muscle strength of the lower extremity in my model by using measured maximum joint strengths. Therefore, I have measured the maximum isometric joint strength of the hip, knee and ankle joint (as torque in Nm) using a CON-TREX WS Dynamometer.

Next, I would like to apply the measured torques to the model and then scale the muscle strength so that for the applied load the muscles which are working against that load show 100% activity. So, for example I would apply a knee flexion moment to the model in order to scale the knee extensor muscles which now need to counter this external flexion moment with an extension moment around the knee joint.

However, when try to apply the load to the model, the model appears to be way too strong. When giving a knee flexion moment of 500 Nm – which is way above my measured values and values from the literature – the model is still not at 100% activity. I am even using HillType-Muscles which, in my opinion, should result in a weaker model compared to Simple Muscles since the full strength of the muscles is not available if the muscle length is not optimal.

I have tried two ways to apply the load to the model. For both approaches I included a file with the following code in the EnvironmentModel so that the torque is included in the inverse dynamics study.

  1. I tried to apply an AnyForce directly to the knee joint measure:
 AnyForce Knee = {
  
  // Here I used two different ways because I was not sure which one is the correct measure for the knee joint but the result is the same
  
  //  AnyKinMeasure &knee = Main.HumanModel.BodyModel.Right.Leg.Jnt.Knee;
  // F = -{500};
   
   
  AnyKinMeasure &knee = Main.HumanModel.BodyModel.Interface.Right.KneeFlexion.KneeRot;
  
  F = {500,0,0};
  
  viewForce = {
    Visible = On;
    ScaleToView = Off;
    AppliedForceColor = {1, 0.1, 0.1}; 
    ForceScale = 0.005;
  }; 
}
  1. I created my own rotational measure between the thigh and the shank and applied the AnyForce to this new measure:
AnyForce Knee = {
  
  AnyKinRotational Knee_Measure = {
    
    AnyRefNode &thigh = Main.HumanModel.BodyModel.Right.Leg.Seg.Thigh.KneeJoint.RotNode;
    AnyRefNode &shank = Main.HumanModel.BodyModel.Right.Leg.Seg.Shank.KneeJoint.RotNode;
    Type = RotAxesAngles; 
  };
  
  F = {-500,0,0};
  
  viewForce = {
    Visible = On;
    ScaleToView = Off;
    AppliedForceColor = {1, 0.1, 0.1}; 
    ForceScale = 0.005;
  };
};

For both approaches the model can handle 500 Nm without reaching 100% muscle activity.

Now my question is: Is the model really this strong or did I not implement the torque in the correct way?

Thank you in advance for your help!

Best regards,
Lukas

AnyBody version: 7.3.4.8518
AMMR version: 2.3.4

2

Hi Lukas,

Welcome to the AnyScrip forum!

Your implementation seems ok! If you can see the muscle activation changing according to the flexion torque, then it should be working alright :slight_smile:

This brings me to the next part. Yea, the model is potentially too strong. I believe the muscle strength comes from the mechanical properties of the muscle (typically extracted from cadavers) and this is possibly a lot more than the usable capacity that a subject can exert. You are right about using the Hill muscle model that it should be weaker than the simple muscle model due to the force-length relationship in the Hill model. Another aspect that could influence the strength of the model is the knee joint axis of the model compared to the subject. This would change the moment arms of the muscle around the knee.

Another point that you can check is that the model is constrained according to the measurements you made.

I hope this helps.

Best regards
Dave

3

Hi Dave,

Thanks for the reply.

About the model strength: Are there any studies that compare the strength of the model - especially the lower extremity - to real subjects? In my experience so far, the model is more on the weaker side. For example, sometimes you get an overload in the lower extremity when using MoCap models, even if the recorded motion represents everyday movements that can easily be performed by the subject. However, I did not find any reliable values that compare the model strength to that of real subjects.

As for the model constraints, the model is held at the pelvis, which should represent the situation of the subject sitting next to the dynamometer. If I am not mistaken, this constraint should not affect the load on the knee joint, since the load is applied directly to the knee joint measure and no further constraints are applied to the thigh or shank segment?

Best regards,
Lukas

4

Hi Lukas,

I can refer you to a couple of studies that could be relevant to you.
https://doi.org/10.1177/0954411916659894

I think the approach you are using seems alright.

Best regards
Dave

5

Hi Dave,

thank you for the literature tip.

Best regards,
Lukas

6

This topic was automatically closed 125 days after the last reply. New replies are no longer allowed.