Hello, I am a new GGEMS user, trying to simulate a cone beam CT (CBCT) system, but am not finding any examples in the documentation about including additional post-source elements (e.g., bowtie and filter) and pre-detector elements (e.g. anti-scatter grid) in the beam path, in addition to the usual CBCT subject (i.e., the digital phantom). The naive option would be to craft a giant phantom grid to include both the phantom and these other elements, which I do not want to do for reasons of computation and memory efficiency.

My understanding is that this can be avoided by using what the documentation calls "navigators" (voxelized or meshed). My questions are,

(1) Is this understanding correct? Is this the intended and preferred way of including additional objects in the beam path?

(2) Can any number of navigators be added, and can the two types be mixed in arbitrary combinations? Could I have for example, in addition to the phantom, a voxelized navigator for the bowtie, a meshed navigator for the pre-bowtie filter layer, and another meshed navigator for the anti-scatter grid?

(3) The documentation mentions something called a "voxelized source". Is this the same as a voxelized navigator? And when providing a voxelized source, is the .mhd file describing a bowtie, or something else?

Hi,

No, you don't need to create one large volume.

(1) Yes — if you want to add a new element to the beam path, you should include it as a navigator (a volume), either meshed or voxelized. GGEMS does not support analytical navigators.

(2) You can add as many navigators as you want, as long as they do not overlap in space. You can also freely combine voxelized and meshed navigators — they are completely independent. So in the example you mentioned (bowtie, filter layer, anti-scatter grid), yes, this setup is fully possible.

(3) In CBCT you do not need a voxelized source. This feature was mainly developed for future GGEMS versions that will support PET and SPECT simulations. It is intended to model photons emitted inside a volume. In a voxelized source, photons are emitted in 360 degrees, so it cannot be used to represent a directional X-ray tube source, if that's what you were considering.

Best regards,
Didier

Thanks a lot! When you say that

GGEMS does not support analytical navigators.

does that mean that the geometric shape generators, like,

Code Select Expand

box = GGEMSBox(24.0, 36.0, 56.0, 'mm')
box.set_position(-70.0, -30.0, 10.0, 'mm')
box.set_label_value(1)
box.set_material('Water')
do not define a navigators? Or are these meshed  navigators in disguise?

No, these are not navigators. I created this tool to automatically generate test volumes, which are then used as voxelised navigators. It's mainly to create phantoms quickly. In the examples, you'll see how I use them.