 |
 |
Hi, You are right that the matrices operate on coordinates in mm. However, the origin is at the corner of the image. If you have your voxels set to 1mm each and you are using ANALYZE then that should mean that the voxel coordinates are equivalent (as long as you start at zero, not one). You can also use the tool img2imgcoord to convert coordinates which might also help you understand what is going on. All the best, Mark On 10 Mar 2010, at 20:24, Ashley Anderson wrote: > I am using FLIRT to register ANALYZE images, but I am left confused > by the > output matrices. The output images look good, so I know the > registration is > successful, but I can't figure out what the shifts in the > registration matrices refer > to (units). Reading past mailing lists, I gather the shifts are > reported in mm. True > units are not important for this project, so I have the pixel size > set as 1mm^3 in > the ANALYZE headers. I hoped this would produce output matrices with > translations in # of pixels, but it doesn't seem to correspond > properly. Is there a > way to get these translations simply in # of pixels? > > Thanks in advance to anyone with suggestions! > > Ashley >