Increased isometric strengths in gait2392_sim

[Tim Dorn]

Hi,

Comparing the models gait2392_simbody.osim, and that which it is referenced to (Delp 1990), I have noticed that the maximum isometric force of all the muscles in the OpenSim model have been unevenly increased by factors between 1.5 - 3. The increases would greatly affect the distribution of muscle force estimates from static / dynamic optimization / CMC / NMT, so I was wondering if there was any justification for this increase.

Cheers,

Tim

[Sam Hamner]

Hi Tim,

I did a little digging to see why these values were different and what effect that might have on muscle force estimations.

There were some notes that were not transferred to the gait2392 model. The comment states "Includes the 43 muscles specified in the Delp leg model and the 6 lumbar muscles included in Anderson's gait model. Peak isometric muscle forces of Delp scaled upward based on joint moment-angle data of healthy young males as done by Anderson (1999) and Carhart (2000, reported in Yamaguchi text)."

To further elaborate, the Delp model, which uses strengths based primarily on cadaver muscle cross-sections, is somewhat weak - and this was observed by both Anderson and Carhart. The the maximum contraction forces were scaled to better reflect Anderson and Pandy's model and the joint torque-angle relationships. This method of "strength scaling" is outlined in Anderson and Pandy (1999) and the Yamaguchi text.

I have attached a PDF that includes the isometric muscle forces from Gait2392, Delp1990, and Carhart2000, along with the scale factors. We tried to do this somewhat uniformly for all muscles at a joint, but bi-articular muscles are an exception as they span two joints. So in the end, it is not a strict uniform scaling, and in some cases Anderson and Pandy's muscle strength parameters were used when each model had the same muscle. (http://wiki.simtk.org/opensim/Gait2392_ ... tricForces)

Finally, I did a comparison of CMC results using the "scaled" Gait2392 and isometric forces from Delp (1990). I have also attached these plots in a PDF. The muscles activations predicted by CMC were not significantly different between the two sets of isometric muscle force. Therefore, I would not expect these increases would greatly affect the distribution of your muscle force estimates, nor your interpretation of those results. (http://wiki.simtk.org/opensim/Gait2392_ ... ResultsCMC)

Here are the references I mentioned:

Delp, "Surgery Simulation: A computer graphics system to analyze and design musculoskeletal reconstructions of the lower extremity," Ph.D. Dissertation, Stanford University, 1990.

Anderson FC and Pandy MG. A dynamic optimization solution for vertical jumping in three dimensions. Computer Methods in Biomechanics and Biomedical Engineering 2:201-231, 1999.

Yamaguchi, G. T. Dynamic Modeling of Musculoskeletal Motion: A Vectorized Approach for Biomechanical Analysis in Three Dimensions, Kluwer Academic Publishing, 2001.

Carhart, M. R. "Biomechanical Analysis of Compensatory Stepping: Implications for Paraplegics Standing Via FNS.," Ph.D Dissertation, Arizona State University, 2000.

Cheers,
Sam

PS. If the PDF links don't work, I have created a page on the wiki: http://wiki.simtk.org/opensim/Gait2392_ ... scle_Force

[Sam Hamner]

sorry, the attachments didn't work. here they are again.

(1) http://wiki.simtk.org/opensim/Gait2392_ ... Forces.pdf

(2) http://wiki.simtk.org/opensim/Gait2392_ ... ltsCMC.pdf

[Margaret Van Horn]

Does anybody know when the Fmax was increased in this model, was the passive force property adjusted also so that the active and passive force relationships remained the same despite increased max isometric force? If so, how did they adjust the passive force property so that it scaled with the increase in Fmax? Thanks for the help!