Understanding intervertebral disc reaction forces

[Riza Bayoglu]

Hi,

Following the tutorial, I predicted vertebral load estimates in the male model and was able to replicate exact same results. I am, however, having difficulty when interpreting disc reaction forces. Do I need to use values from the parent or child components when reporting compression and shear forces? My feeling is that I should use values from the parent components.

For example for the L5/S1 joint, child and parent component results are:

child

L5_S1_IVDjnt_on_lumbar5_in_lumbar5_fy 576.4
L5_S1_IVDjnt_on_lumbar5_in_lumbar5_fx -141.35
L5_S1_IVDjnt_on_lumbar5_in_lumbar5_fz 0.0

parent

L5_S1_IVDjnt_on_sacrum_in_sacrum_fy -588.77
L5_S1_IVDjnt_on_sacrum_in_sacrum_fx -74.6
L5_S1_IVDjnt_on_sacrum_in_sacrum_fz -0.0

Thank you.
Riza

[Jacob J. Banks]

Yes. The JRA analysis is where you get the load estimates you (I assume) are interested in. Fy should be Compression, Fx should be AP shear, and Fz should be ML shear.
The parent body should be the superior (as opposed to inferior) body relative to the joint. The Child the inferior.

It really doesn't matter which you use, so long as you report it. Shear will be a little more sensitive to the coordinate system, but you'll find the 3D resultant force to be the same.

In Anderson's lab, we often report a vertebral load, which is the average between inferior and superior joints (so for L5 vertebrae, it would include the Parent of the L5/S1 and the child of the L4/L5).

So, back to your original question... it doesn't matter, so long as you report it. I wouldn't worry about 'signs' either, so long as you report/define them.
Jake Banks
PS: Sorry about the delayed response, just seeing this... my PC was impacted by CloudStrike last Friday.

[Riza Bayoglu]

Hi Jake,

Thanks for the helpful comments and your patience. I was traveling past few days and wanted to explore some things before I comment here again. I verify that the resultant joint forces are the same, the differences are purely due using different coordinate systems. I am attaching the results I extracted from your model. Could you please help me understand the directions of the shear forces? I am now thinking of reporting results from the child object. In this case, is it correct to say that AP shear force is directed posteriorly at the T1/T2 joint and anteriorly at the L1/L2 joint? For the ML shear force, is it directed posteriorly at the T1/T2 and anteriorly at the T2/T3 joints? Or do I need to flip the signs and directions?

Best,
Riza

joint forces.PNG (49.18 KiB) Viewed 990 times

[Jacob J. Banks]

Riza,
This always makes me think for minute... so, I could off on my logic without thinking about it too much (and not looking at any notes I have)... but I believe (from https://opensim.stanford.edu/downloads/ ... action.pdf and working off from FBD and the sign on the Compression Forces)

The parent just goes off from the coordinate system you see for that body, so:
+ Tension
+ Anterior
+ Right

Then for the Child, it is just the opposite:
+ = Compression
+ = Posterior AP shear
+ = Left ML shear

But this is very confusing to think about and even I end up going back/forth. This is why 'Resultant Shear' is a nice fall back! Further, shear forces are so dependent on muscle orientations etc., (Kingma Ref here) that I never feel overly confident in reporting them.

I'll let you know if I think of a better way to explain or anything that may contradict what I am thinking. Have a good one,
Jake Banks

[Riza Bayoglu]

Hi Jake,

I still have difficulty understanding the sign convention for the shear forces. I agree that reporting the resultant shear forces is a safe choice. I'll proceed with the resultant forces.

On a different topic, I am looking at the musculotendon force output (NMB_MuscleAnalysis_TendonForce.sto file).
I created two models to compare tendon forces. The first one is the standard male model, and the second one is the body mass scaled version of the first model. So, the only difference is the total body mass and the mass and inertia distributions across the body segments. The scaled model is +30 kg heavier. What I noticed was that although compressive joint forces and muscle activations are significantly higher in the scaled model (as expected), tendon forces (also active and passive forces) are almost identical (differences are less than 0.2 N if not identical). I don't quite understand this. Do you have a clue what might have caused this behavior?

Thanks,
Riza

[Jacob J. Banks]

Sorry, I have been at the American Society for Biomechanics conference.

How are you computing the Force variables?
The Tendon force should be = Muscle force. If they are not, I would trust the Muscle forces from SOpt.
There are no passive forces in this model, as SOpt does not consider Parrell elastic forces (https://opensimconfluence.atlassian.net ... tion+Works)
Jake Banks

[Riza Bayoglu]

I am calculating muscle forces using static optimization, this process extracts all the force output files. I attached my setup file for the static optimization and results from the male model. Although the OpenSim references says there is no passive fiber force, I see them in the results. I checked the setup file, but did not see any options around including passive forces. Why is then passive fiber forces are calculated?

I verified that;

Tendon force = Active fiber force along tendon + Passive force along tendon
Fiber force = Active fiber force + Passive fiber force

Tendon force is not equal to fiber force, fiber force is slightly larger. This is expected since tendon force is a projection of fiber force along the tendon line-of-action.

Upon looking at the static optimization force data (NMB_StaticOptimization_force.sto), I see these forces are very different than tendon forces. Which one is correct?

[Jacob J. Banks]

viewtopicPhpbb.php?f=91&t=7057&p=18635&start=0&view=
The above link should answer your question better than I did. In brief, all you can get from SOpt is muscle force. The tendon forces that are "calculated" are not being considered in SOpt, and that is why they do not give with one another etc.
Jake Banks

[Riza Bayoglu]

This was very useful. I confirm that muscle forces were higher in the scaled model (higher body mass). I also checked that joint reaction analysis uses the muscle forces from the same file (NMB_StaticOptimization_force.sto), so joint reactions should also be correct. Thanks, Jake! I appreciate your help.