 |
 |
Hi again, to see if this has any positive effect just add —userefmask=0 to your fnirt command line, alternatively in the config file. If you have a lot of signal from extracerebral tissue, in particular marrow, this is not likely to improve matters. But if your structurals have been acquired with fat-sat and you cannot see any bright areas outside of, but close to, the brain surface it should work well. Jesper On 14 Apr 2014, at 13:19, Andreas Lidström <[log in to unmask]<mailto:[log in to unmask]>> wrote: Thanks, but just but be clear; when is the refmask used in the process and should I simply not use one or specify another mask? 2014-04-14 11:46 GMT+02:00 Jesper Andersson <[log in to unmask]<mailto:[log in to unmask]>>: Hi Andreas, this is typically caused by warps extending to, and beyond, the edge of the brain where there is little or no information for the last steps of the structural->standard brain together with a so-so structural->functional registration (typically from lack of fieldmap). What you can try is to edit the fnirt config file so that you don’t use the refmask. If your T1 is a good match for the structural in terms of extracerebral tissue (i.e. if it has been acquired with fat-sat) it might help informing the warps at the edge of the brain. Jesper On 14 Apr 2014, at 10:28, Andreas Lidström <[log in to unmask]<mailto:[log in to unmask]>> wrote: > Hello, I have been running two step registration of fmri into MNI space and the final outcome is images with strange flaring radiations out from the area matching the standard template. This is worst in the parietal regions (see attached image). What can be the cause of this? > > The registration between my T1 and the template looks good on the report page and my functional to T1 looks reasonable. > > Is this maby a masking problem or a failure of BET on the functional data? > <fMRI wings.png>