Complicated motion capture, 6 DOF of femur, tibia, and patella.

Trying to figure out how to transform marker data points (x, y, z) into joint angle information (6 DOF) of the knee joint. In my model there are 3 markers attached to each bone (femur, patella, tibia) via 6 inch surgical pins (the pins are screwed into cadaver bone and the markers are attached to the pins). I would like to avoid having to use Computed tomography as well.

Anyone know of a good method to go about doing this? Many of my colleagues are only used to dealing with surface markers and gross motion, not 6 DOF of the knee (both tibiofemoral and patellofemoral).

[jo woo]

This (<http://www.ncbi.nlm.nih.gov/pubmed/20170853>) is very similar to what I am planning to do except I am trying to avoid 3d reconstruction as I do not have access/funding to obtain. I would like to come up with transformation matrices to calculate joint motion with 6DOF using marker position and plain film radiographs that show marker position in relation to bone position. Any ideas?

[Star Strider]

I would search the robotics literature as well. If I remember correctly from my brief (and ancient) exposure to 3D robotics (in the context of rehabilitation medicine), the roboticists have a relatively well-established and sophisticated way of dealing with these problems. They were in the context of control though, so I have no idea if those would apply to what you are doing. Beyond that, I can offer no suggestions.

Good luck!