simtk.org

Hi, fellow workshop-attendees,

I modified the gait2392 model so now the shank segment is made of two bodies (proximal and distal tibia) connected via a weld joint at the tibial midshaft. I aim to use joint reaction force plugin to estimate forces and moments at the midshaft. Anatomically, some muscles have large attachment area and attach to both proximal and distal part of the tibia (e.g., flexor digitorum longus) and I would like to reflect this in the model.

What would be the best way to split the origin of a muscle into two while keeping their insertion at the same location (e.g., creating flex_dig1_r and flex_dig2_r)?

1) Muscle path geometry
I am thinking to move the origin of each new muscle to the center of the corresponding attachment area (either proximal or distal from the midshaft). From here I would go straight to the insertion of the original muscle or to first via point of the original muscle.

2) Muscle parameters
I am thinking to adjust only the maximum isometric force of the new muscles as it would be difficult to validate changes in fiber length or pennation angle. Does it sound right to (1) adjust the maximum isometric force proportionally to muscle attachment area while (2) the sum of the new musce forces equals the force of the original muscle?

Thank you in advance for your suggestions!
Best, Martin

I am thinking to move the origin of each new muscle to the center of the corresponding attachment area (either proximal or distal from the midshaft). From here I would go straight to the insertion of the original muscle or to first via point of the original muscle.

Seems like a reasonable strategy; however, you might want to avoid large differences in the moment arms as that could result in large changes in muscle activity, insufficient moment-generating capacity, etc.

I am thinking to adjust only the maximum isometric force of the new muscles as it would be difficult to validate changes in fiber length or pennation angle. Does it sound right to (1) adjust the maximum isometric force proportionally to muscle attachment area while (2) the sum of the new musce forces equals the force of the original muscle?

You might also want to change optimal fiber length and tendon slack length since the two paths will be different lengths. You could try doing something like what the muscles do during scaling:
1. Put the model in its default pose.
2. Measure the length of the original muscle path.
3. Measure the length of the new muscle path.
4. Set the optimal fiber length of the new muscle to the optimal fiber length of the original muscle multiplied by the ratio of the two path lengths (new/original); similarly for the tendon slack length.

Thank you, Tom.

I will check the effect of path geometry on muscle moment arms and follow your suggestion on scaling the optimal fiber length and tendon slack length.

Martin