computer models of biomechanics: muscles

[gusl]

I've seen lots of 3D models of the human skeleton. Can we model muscles on top of this as rubber bands stuck to different parts of the bones that shrink when tensed?

My mom is a physical therapist and believes that there would be a lot of money on this. Such a system could, for instance, illustrate how different stretching exercises have different consequences, and the different ways in which people can compensate for a given muscle shortening.

Properties of human muscles:
* muscles stretch
* the base-level length of a muscle can stretch (i.e. stretching for the second time is easier than for the first) rather quickly, and will shrink slowly if not stretched in a long time.

Things I need to figure out if I am to do such a project:
* how to find and use graphics APIs (preferably for Common Lisp!)

Once I get playing with the 3D models, I imagine that figuring out the mechanics itself will be fun and easy, although maybe I will need data about real people.

Comments

[jcreed.livejournal.com]

I feel it's absolutely certain people have done at least something like this in modern animation.

[jcreed.livejournal.com]

Whether it's available to anyone else is a separate question.

[gustavolacerda.livejournal.com]

I agree. So the question is why this isn't widespread in physical therapy circles?

[jcreed.livejournal.com]

Yeah, this may well be a good niche where one group of people has effectively solved a problem, and all one needs to do is the work of "exporting" it...

[smandal.livejournal.com]

Two complicating factors come to mind:

* Muscles attach over macroscopic areas, with different density of muscle fibers at different points.

* The geometry of muscle, and therefore the mechanics, change as muscles are trained -- fibers grow and develop different contracting and elastic properties. Furthemore, fibers do not train uniformly for a given exercise regimen.

[gustavolacerda.livejournal.com]

But can't you get a pretty good approximation by modeling each muscle or fiber as a string with a single number representing its elasticity?

[smandal.livejournal.com]

I don't think so, because the size of the attachment regions are of the same order of magnitude as the size of the muscles themselves. At least, for the large muscles. For the smaller muscles you might be able to get away with it, but then you won't capture the torquing action some of the sheet-like muscles exert.

[simonfunk.livejournal.com]

Jurrasic park (and presumably Kong, etc.) was rendered this way. I assume by now there are standard add-ons to do this for Maya, etc.

[simonfunk.livejournal.com]

http://www.newscientisttech.com/article/dn8953.html

[gustavolacerda.livejournal.com]

Cool. Thanks.