KAHIBARO
30.4. Gamma Camera Geometry
Table of Contents
Detector head
A gamma camera detector head is the part of the system that directly views the patient and detects incoming gamma photons from the radionuclide. In GATE you represent the detector head as a set of geometric volumes that together approximate this physical structure. For beginners it is useful to think of the head as a flat rectangular assembly placed close to the source or phantom, with its normal aligned with the direction of incident photons.
In a basic SPECT simulation the detector head sits behind the collimator. You typically build the collimator geometry first, then place the head immediately behind it along the main axis of detection. For a simple setup the head can be a single box volume that acts as a parent for more detailed components such as the scintillation crystal, light guide, and photodetectors. In more advanced models these internal components become separate child volumes, each with its own material and role in the simulation.
The key geometric parameters of the head are its transverse size, which determines the field of view, and its thickness, which you choose large enough to contain the crystal and any backing structures. In SPECT fundamentals you usually align the head so that its sensitive surface is planar and parallel to the collimator face, for example in the XY plane with photons travelling along the Z axis.
From a simulation perspective the detector head volume is often used as a convenient parent for attaching actors that record detector-related information. Although the main interactions of interest occur in the scintillation crystal, the head volume provides a logical container that keeps all detector components grouped together and simplifies later transformations such as rotations during SPECT acquisitions.
Crystal
The scintillation crystal is the core sensitive element of a gamma camera. In simulation you must specify its geometry and material carefully, because they have a direct impact on detection efficiency, energy spectrum, and spatial response. For low and medium energy SPECT a common choice is a NaI(Tl) crystal, modeled in GATE with the corresponding Geant4 material. Typical dimensions are a few millimeters thickness and a transverse area similar to the field of view of the detector head.
In most beginner SPECT simulations, the crystal is represented as a single rectangular box placed just behind the collimator. You define its width and height to match the useful imaging area and its thickness according to the application energy range. Thicker crystals increase detection efficiency but may degrade intrinsic spatial resolution. In GATE, the crystal volume is usually a direct child of the detector head volume. This makes it easier to handle motions of the entire head, because the crystal automatically follows the parent transformations.
The position of the crystal relative to the collimator is important. There should be a small but realistic gap representing the light guide or mechanical structures, but for many educational simulations you can place the crystal very close to the back of the collimator without modeling all intermediate layers. You only need to ensure that there is no geometric overlap between the crystal and the collimator volumes, since overlaps can cause problems in Geant4 based simulations.
The crystal material determines how gamma photons interact. In a fundamental SPECT geometry you typically focus on photoelectric absorption and Compton scattering processes within NaI(Tl) at around 140 keV for Tc 99m. The amount of energy deposited in the crystal volume is what you later convert into detector signals through the digitizer chain. For this reason you usually attach actors to the crystal or to a thin layer within the crystal to record hits, energy deposition, and timing.
In more advanced models you can subdivide the crystal into smaller volumes to approximate position sensitive detection, for instance an array of pixels. However, for an introductory SPECT gamma camera geometry a monolithic crystal is sufficient to demonstrate the main imaging concepts and to generate projection data.
In a basic gamma camera simulation the scintillation crystal is the main scoring volume for gamma detection. Most detector actors and digitizers should be configured to use hits and energy deposition in the crystal volume, not in the whole detector head.
Shielding
Around the detector head and crystal you normally add shielding material to reduce background counts from photons that arrive from outside the useful field of view. In clinical gamma cameras, high density lead is used extensively around the sides and back of the detector. In GATE you model this shielding as additional volumes that surround the head, leaving only the front face of the collimator exposed to the patient.
A simple way to implement shielding in a fundamental geometry is to create a larger box of lead that encloses the detector head and then cut out the area where the collimator and crystal face the source. You can approximate this by building separate lead slabs on the sides, top, bottom, and back of the head while leaving the patient facing side open. The exact thickness of the shielding can be adjusted according to the photon energy you simulate, but it must be thick enough that few photons penetrate it compared with those passing through the collimator.
The shielding has two main effects in the simulation. First, it absorbs photons that would otherwise reach the crystal from undesired angles, which helps you model a more realistic background level in the detector. Second, it slightly increases the number of scattered photons in the vicinity of the camera, because some photons may Compton scatter in the lead before escaping or being absorbed. When you analyze projection data you will see that shielding helps limit counts from out of field activity and improves image contrast.
In your geometry hierarchy the shielding volumes are usually siblings of the detector head volume, all placed as children of the world volume. They are positioned so that they tightly surround the detector head, but again without any geometric overlaps. It is often convenient to give them clear names, such as side shielding and back shielding, so that you can visualize them easily and, if needed, apply specific material or thickness changes for sensitivity studies.
Although the internal electronics and support structure are not normally included in an introductory SPECT gamma camera simulation, adding basic shielding geometry is important to capture one of the essential characteristics of real gamma cameras, which is the strong directional sensitivity enforced together by the collimator and the surrounding lead.
Views: 9