Hi All,

I kept forgetting to bring this up, but I think we are missing an important piece of the transformation matrices, that being what the coordinates actually mean. For example (-24, -6, -16) might mean the left amygdala in Talairach space. Or it could mean the right parahippocampal gyrus, even in Talairach space. Does the -24 mean 24L? 24R? 24I?

It would be sufficient (and probably prudent) to say that we follow the coordinate definition from NIfTI, which describes coordinates as being L->R (negative to positive), P->A (- to +), I->S (- to +). Following this convention, the (-24, -6, -16) example would mean 24L, 6P, 16I, which would be in the left amygdala (when using the Talairach template space).

Does that seem reasonable to everyone? If so, it should be in the documentation. Maybe it is and I have just completely missed it. If so, uh, never mind...

- rick

Rick,

You are correct, we never explicitly stated the orientations (and the units?) of the coordinates and this is important. In fact, this was recently pointed out during the CIFTI discussions. CIFTI will follow the NIFTI convention and so should GIFTI.

If there is no disagreement, I will update the GIFTI document to state that positive X is right, positive Y is anterior, and positive Z is superior. In addition, the units of the coordinates are millimeters.

John

Thanks, John.

- rick

An updated version of the GIFTI Surface Format document has been uploaded. Section 13.1 has been updated to indicate the orientations of coordinates and that coordinates are in millimeters.