[Gate-users] Non-circular orbit in SPECT acquisition

[Hesz Gabor]

Hello,
if nothing else works, you can create a rotated phantom and source for 
each angle, crop air from the sides and do the simulation separately for 
each projection angle. This way the detector is always parallel to the 
volume's edge and won't overlap with it. (You can use 2 or 4 detector 
heads simultaneously for each view to speed up the simulations.) Make 
sure the cropped volumes are translated correctly.

Gabor


2017-08-23 13:41 keltezéssel, Gustavo COSTA írta:
> /Hello GATErs,/
>
> /I'm trying to mimic a non-circular tomographic acquisition in GATE 
> meaning that I'll use the minimal detector/phantom distances that were 
> retrieved from real acquisitions./
> /I have 2 problems:/
>
> /1 - Model different distances to each detector head./
> /2 - Since I'm using a voxelized phantom from a DICOM image, it is a 
> cubic image and I get overlaps between the detector and the corner of 
> the voxel image made of air (because the detector rotates closely to 
> the phantom)./
> /Any ideas how I could simulate that ?/
> /You can see below what I have already tested and the errors I got./
> /Thanks in advance/