KAHIBARO
26.2. PET Scanner Components
Table of Contents
Detector crystals
A PET scanner is built from many small detector crystals that convert incoming 511 keV photons into measurable electrical signals. In GATE you represent these crystals as individual volumes with a specific material, size, and position. The material is usually a scintillator such as LYSO or BGO that produces visible photons when a gamma photon deposits energy.
Each crystal has three key geometric dimensions: width, height, and depth. In PET, the depth of the crystal along the radial direction is particularly important because it influences detection efficiency and depth of interaction effects. Crystals are typically a few millimeters wide and several millimeters to a few centimeters long. In a simulation you treat each crystal as an individual sensitive volume where hits are recorded and later processed by the digitizer chain.
Crystals play two main roles in a PET model. First, they define where gamma photons can be detected, because only interactions inside these crystal volumes contribute to the PET signal. Second, they define the spatial sampling of the scanner. The crystal cross section determines the intrinsic spatial resolution and the granularity of the detector response. When you design a PET geometry in GATE, you will therefore specify not just the crystal material, but also the exact crystal pitch and arrangement to match a real or hypothetical scanner.
In simulation output, each crystal is often identified by one or more indices such as crystal ID or its indices inside a crystal array. These indices allow you to map detected events to physical positions on the detector, which is essential for later steps like forming lines of response and reconstructing PET images.
Important: In PET simulations each detector crystal must be a separate volume with a unique identity. This allows the digitizer to group hits per crystal and to create singles and coincidences correctly.
Rings
Detector crystals are not placed randomly around the patient, but organized into one or more rings that surround the field of view. A PET ring is a circular arrangement of detectors around the patient axis. This cylindrical geometry is what allows PET to detect pairs of photons in coincidence across the body and to form lines of response.
In GATE you do not usually place every crystal by hand on a circle. Instead, you define a basic detector unit, for example a crystal or a block of crystals, and then repeat it around a ring using parameterized or repeated volumes. Two main parameters describe a PET ring: the ring radius or diameter, and the axial length. The radius determines the distance of the crystals from the center of the scanner and strongly influences sensitivity and spatial resolution. The axial length, that is the coverage along the patient axis, determines how much of the body you can image at once.
A typical PET scanner consists of several contiguous rings stacked along the axial direction. In simulation, you can represent this by repeating your base detector module along the axial axis to build a full 3D cylindrical detector. The combination of azimuthal repetition around the ring and axial repetition along the scanner length gives you the complete detector envelope.
Rings are also important for detector identification. Many scanners use a hierarchical ID structure where each detected event is associated with a ring index and a crystal index within that ring. You will use these indices when constructing lines of response or analyzing spatial patterns, for example to study sensitivity as a function of axial position.
Key rule: The ring radius and axial coverage must be consistent with your intended scanner design. Changing these values changes sensitivity, field of view, and spatial resolution in a PET simulation.
Modules
Between individual crystals and the full ring there is an intermediate structural level: the detector module. A module is a logical group of crystals and often corresponds to a physical component such as a detector block, a tile, or a block coupled to a readout device like a PMT or SiPM array.
In GATE a module is typically a volume that contains a regular array of crystals. You define the module dimensions so that the crystals fit inside with the correct pitch and possible gaps. Then you place many identical modules around the ring and along the axial direction. This modular design offers two advantages. It reflects how real scanners are built from repeated hardware components, and it makes the geometry definition much more compact and flexible. If you change the crystal size or layout inside a module, the whole scanner can update automatically through the repetition of that module.
Modules are also central for organizing detector identifiers. A complete detector ID may include indices for ring, module, and crystal within the module. This hierarchical structure lets you translate between physical coordinates and channel numbers in the readout electronics. In the digitizer and in the analysis of singles and coincidences, you will often use these identifiers to map events to specific hardware elements or to group signals by module.
Different PET designs use different module layouts. Some use small blocks with many thin crystals, others use long crystals grouped into larger panels. In simulation, you can explore these configurations simply by changing the module geometry and repetition pattern, while keeping the overall ring concept constant.
Remember: Use a clear hierarchy crystal → module → ring in your PET geometry. This hierarchy simplifies geometry construction, detector identification, and later analysis of PET events.
Views: 9