Motions of the left side of the body

[Marco Ganzetti]

Dear members,
in order to describe the angular kinematics of the left side of the body (in particular the upper body) which is the best method recommended?

Whereas I use the global coordinate system defined as: Y vertical, pointing upward, X pointing from left to right, Z pointing backward.
I tried to reverse only the sign for the X-coordinate of each raw marker (3D data expressed in the global coordinate system).
Doing so, I can avoid left-handed coordinate systems or changes of definition for each anatomical reference system (using the same matlab program for both sides, instead of define a similar reference system/program for the left side, too).

Repeating the same procedure for the Z-coordinate.... it seams that I get the same result?

Any ideas would be very much appreciated.

Thanks,

Marco

[Ton van den Bogert]

Re: Motions of the left side of the body

Marco,

What you did is perfectly fine. I have sometimes used the same trick to avoid having to write separate code to process data from the other side of the body.

Essentially, by inverting the X values, you are looking at your subject through a mirror. So the left side looks like the right side, and you can process the left side data as if it were the right side.

If you invert the Z values instead, you are also looking at your subject through a mirror. This time, the mirror is in the frontal plane, so it switched front and back instead of right and left. But it is still a mirror, so your left side looks like the right side. So it makes perfect sense that this works equally well!

Ton van den Bogert

[Marco Ganzetti]

Re: Motions of the left side of the body

So it's as I thought, I can use interchangeably both of them (changing X or Z sign) with the same outcome.
But, I was wondering if this is also right with the subject rotating 180°....in other words if the subject change direction during the acquisition (facing back).

Thanks,

Marco

[Ton van den Bogert]

Re: Motions of the left side of the body

The orientation of the subject should not matter if you use a 3D model to calculate joint angles.

This would be easy to verify: invert all horizontal coordinates (X and Z) to simulate a 180 deg rotation about the Y axis, and run the data through your software again. The results should be the same as before.

Ton