Hi,
There are some previous postings on the flirt matrix conventions
in the FSL list. However, in general we don't recommend trying
to do things with the matrix itself if you can avoid it, as it can be
time-consuming and annoying to get all the details of the
conventions right, but it depends what you need this for.
I don't believe it matters what range the priors are in fast as
I think it normalises them within the code, but to be on the safe
side you can make them the same range of values as those
in $FSLDIR/data/standard/tissuepriors/
which is from 0 to 255 (representing 0 to 1.0 for historical reasons).
It doesn't matter what the datatype is.
All the best,
Mark
On 11 Feb 2010, at 18:44, Torsten Rohlfing wrote:
> Hi --
>
> Is there any information (e.g., on the FSL web site) that describes
> the
> flirt matrix format, in particular how the matrix relates to the image
> coordinate systems? Likewise, is there any documentation on the
> format that
> fast expects tissue prior images to be in (e.g., pixel data type,
> permissible value range)?
>
> Thank you!
> Torsten
>
|