Dear Wolf (et al.),

We use FLIRT routinely on rat brain data for spatial normalisation to our stereotaxic template. To deal with the size issue that might give rise to subtle (or not so subtle!) effects due to any 'human-sized' assumptions when using s/w such as FSL, SPM etc with small animal data, we simply scale up the image header dimensions x10 (when we first transform from scanner to Analyze format). I believe this is a fairly common trick. Coregistration and subsequent processing work just fine.

Hope this helps
adam


"FSL - FMRIB's Software Library" <[log in to unmask]> wrote on 13/06/2005 15:39:34:

> Hi all,
>
> If using flirt on animal brains it was in general suggested to rescale
> the data and fake a human sized brain. If I obtain a transformation
> matrix using this method, how di I nned to modify it to apply the
> transformation on the original data? For example I rescale EPI and HiRes
> Images with a factor of 3, perform flirt, which gives me a
> transformation matrix. Now I would like to apply this transformation
> matrix on an activation map, so what modification are required?
>
> Thank you for all the help!
>
> wolf
>