We designed a partial PET scanner based on a cylinder PET scanner but we could not reconstruct the exact image. Is there a limit to the Castor application that cannot reconstruct a dual PET scanner?



The limit is not in CASToR but is mathematical. If you cannot record enough projection angles of your object, the problem is under-determined and you cannot recover the exact object.



As a user of CASToR, I have extensively use it for 2-head PET reconstruction for experimental and simulation data. But there are many parameters that could go wrong. It needs caution.


Dear CASToR users

I have been using this platform some months with iterative algorithms. Nowadays, I am trying to reconstruct some small spherical sources with FBP2D algorithm. Is there any possibility to achieve this goal with CASToR? I read time ago that this algorithm is not implemented, but just to be sure



Hi Gabriel,

Indeed, it is not implemented has it requires a sinogram representation of the data with invariant sinogram bin size.

In some future, there will be a “sinogram scanner” in CASToR so that any type of PET, SPECT or CT data from a real scanner could be rebinned into a user’s specified sinogram representation.

For the moment, you will have to look somewhere else for FBP2D.