Locked Coordinates with Inverse Kinematics

[Emily Hubbard]

Is anyone familiar with locking coordinates, specifically mtp and subtalar, for scaling and the effect on inverse kinematics? In one case I unlocked the coordinates for inverse kinematics and one foot became malformed. In another case I left the mtp and subtalar coordinates locked, which gave me a much more normal gait.

I'm wondering if anyone knows the degree of detriment to the knee angles and knee forces data as a result of keeping the coordinates locked. Essentially, do you think it's better to lock the coordinates and have a normal gait or to not lock the coordinates and have an abnormal gait?

Emily

[jimmy d]

I'm wondering if anyone knows the degree of detriment to the knee angles and knee forces data as a result of keeping the coordinates locked.

There is no immediate reason to think that there would be a detriment to knee angle or moment estimates. There is not enough resolution in marker based motion capture to get the precision needed to track these coordinates, especially if you only have a few markers on a shoe. Keeping these DoF locked is well within reason.

[Emily Hubbard]

James,

I have done some reading and now have tried just to lock mtp joints, which warrants a normal-looking gait seen here: compared with the unlocked joints here: and

However, I plotted the right knee angles for inverse kinematics, because the right foot is the only foot giving me a problem. The plot is for left and right mtp and subtalar locked, just mtps locked, and nothing locked. The graph shows a variance in knee angles of a few degrees:

What do you think of this difference? Does it seem significant enough that I shouldn't lock joints and should instead use the abnormal simulation?

Thanks for the help.

Emily