KAHIBARO
13.3. Region-Based Cuts
Table of Contents
Detector regions
Region based cuts allow you to control how precisely particles are transported in different parts of your geometry. Instead of using a single global production cut for the whole simulation, you can assign different cuts to logical regions that correspond to specific volumes such as detectors or phantoms. This is essential in medical physics, where small detector structures and large patient volumes coexist in the same simulation.
In GATE, a region is a logical grouping of one or more volumes. Each region can have its own set of production cuts for gammas, electrons, positrons, and possibly other particles, as well as its own transport parameters. The detector system is usually where you need the highest spatial and energy resolution, because you want to reproduce measured spectra, timing, and spatial response as accurately as possible. For that reason, the detector region often uses smaller production cuts than the rest of the simulation.
A typical workflow is to define a region that includes all sensitive detector components, for example scintillation crystals and possibly light guides or immediate surrounding materials. The support structures or shielding might be left in the default region or grouped into another region with more relaxed cuts. In GATE, you attach a region to a logical volume, and this region is inherited by its daughters unless you explicitly attach a different region further down in the hierarchy. This hierarchical behavior lets you set fine control on where detailed physics is needed.
The main idea for detector regions is to balance realism and speed. A small cut in the detector region means that low energy secondary particles will be generated and tracked, which improves the accuracy of quantities like energy deposition patterns, peak shapes in energy spectra, and coincidence timing. However, this also increases the number of steps and the CPU time. For high resolution PET or SPECT detector modeling, it is common to use cuts that are significantly smaller than the crystal size, so that the geometry of the detector is well resolved in the particle transport.
You should define detector regions in situations such as small scintillation crystals, semiconductor detectors, or thin converter layers in a gamma camera. In those cases, a coarse global cut might be larger than the detector thickness, which would suppress important secondary production and distort the detector response. By setting tighter cuts locally, you ensure that the detector physics is accurately simulated without forcing the same level of detail everywhere.
Important rule: Always choose production cuts in the detector region that are small compared to the critical detector dimensions, such as crystal thickness or pixel size, so that detector response is not artificially smoothed or biased.
Phantom regions
Phantom regions group the volumes that represent the patient or other extended objects such as water tanks, anthropomorphic phantoms, or voxelized geometries. The goal in phantom regions is usually accurate dose or energy deposition information, but with less emphasis on resolving very fine structural details than in detector crystals. This often allows you to use larger production cuts than in the detector region, which can significantly reduce computation time while keeping dose accuracy within acceptable limits.
A phantom region is typically attached to the main phantom volume, like a water box or the root of a voxelized patient geometry. All internal voxels or subvolumes then inherit this region, unless you define specialized regions for particular organs or substructures. This is useful when you want higher detail in a small organ, for example the eye or a tumor, but can tolerate coarser transport in the rest of the body.
In dose calculations, production cuts in the phantom region should be chosen with respect to the dose scoring resolution. If you use a dose actor with a given voxel size, you normally select cuts that are not larger than the smallest dimension of the dose voxels. That way, the transport step lengths and the creation of secondary particles are fine enough to represent the spatial distribution of energy deposition within each voxel realistically. If the cuts are too large compared to voxel size, the simulation may miss small scale features, for example the sharp rise and fall of dose near material interfaces.
Another important consideration in phantom regions is the dominant particle type. In photon therapy or nuclear medicine, electrons produced by photon interactions are often responsible for the local energy deposition. In that case, the electron production cut in the phantom region has a strong effect on the spatial accuracy of dose and should be chosen carefully. For proton therapy simulations, you also need to ensure that cuts and transport parameters in the phantom are appropriate for resolving the Bragg peak and the distal fall off.
There is usually no benefit to use very small detector like cuts in the entire phantom region, because this would dramatically increase computation without significantly improving clinically relevant quantities such as mean organ dose. Instead, you adjust phantom region cuts to the level of spatial detail you need for your specific dosimetric or imaging question.
Key guideline: In phantom regions, select production cuts that are compatible with the spatial resolution of your dose or activity scoring grids. Cuts much smaller than the scoring resolution waste CPU time, while cuts larger than voxel dimensions can degrade the accuracy of spatial dose distributions.
Views: 12