 |
 |
Hi, You are right that Prelude is unhappy with disconnected parts of the image. It does some connected component operations to fill small holes in masks and this will prevent it working on disconnected areas. So the solution is simply to make a separate mask for each disconnected area and process them separately through prelude. This is very easy to script and you can easily get the separate components from your mask using cluster - e.g. cluster -i mymask -t 0.5 -o indeximage will create and image called "indeximage" which has a different non-zero value for each separate component in your mask. You can then make separate files using fslmaths with the -thr and -uthr values set appropriately. I hope this helps and sorry for the delay - I just got back from post FSL-course holiday. All the best, Mark Marnix Maas wrote: > > Dear FSL people, > > I’d like to use FSL to correct for the distortions in EPI images of a > phantom consisting of several bottles of liquid. I made field map > images of this phantom consisting of a magnitude and a phase image, > and converted the phase image to radians using fslmaths. Then I passed > the field maps to Prelude to unwrap the phase. > > The result however was that only 1 bottle showed up in the unwrapped > image. I’m guessing that this is because the unwrapping algorithm > stops when it encounters areas of too low intensity in the magnitude > image. > > So my question is: is it possible to avoid this (rather > non-brain-like, admittedly) problem e.g. by passing multiple starting > points to Prelude, or by some masking tricks? > > Thanks in advance, > > Marnix Maas > > Postdoc Physicist > > Netherlands Cancer Institute / > > Antoni van Leeuwenhoek Hospital > > Department of Radiotherapy > > Plesmanlaan 121 > > 1066 CX Amsterdam > > The Netherlands > > Tel: +31 20 512 1729 > > Fax: +31 20 512 1745 > > e-mail: [log in to unmask] <mailto:[log in to unmask]> >