In spm12b it is a batch option. Otherwise, you have to modify spm defaults. I asked this question a few days ago. It seems that the 0.80 percent is high for uneven images that are a product of multichannel coil receivers. Obviously SNR is weaker for uneven signal, but having huge dropouts is not very fancy either.
Stan
On Feb 10, 2014, at 6:47 AM, Micael Andersson UMU <[log in to unmask]> wrote:
> Dear SPM users
>
> We have noticed that the default mask threshold (80% of global mean) is to high for our fMRI-data. It do not cover the whole brain. We have tried to lower it and instead put an explicit mask of GM+WM+CSF. Are there any drawbacks by doing this? If not, why is not the mask threshold an option in the SPM-batch?
>
> --
> Micael Andersson
> Forskningsingenjör / Research engineer
> Umeå centre for Functional Brain Imaging (UFBI)
> Fysiologi, IMB
> Umeå universitet / University of Umeå
|