KAHIBARO
30.2. SPECT Scanner Components
Table of Contents
Collimator
In a SPECT system, the collimator is the first component that photons encounter after leaving the patient. Its main role is to select the direction of incoming gamma rays so that the final image reflects where in the patient the photons originated. Without a collimator, the detector would measure photons from all directions and the image would lose spatial information.
A typical clinical SPECT scanner uses a lead collimator placed directly in front of the scintillation crystal. It is built as a thick plate of lead with many parallel holes. Each hole acts like a simple “acceptor” for photons whose direction is roughly aligned with the hole axis. Photons that approach at larger angles are absorbed in the lead septa, so they never reach the crystal.
You will meet several standard collimator designs in SPECT, for example parallel hole, converging, diverging, and pinhole. In this chapter we focus on their role as a scanner component, not on detailed types. For a parallel hole collimator, the most important geometric parameters are the hole shape, the hole diameter, the septa thickness, and the collimator thickness. These parameters determine the trade off between system sensitivity and spatial resolution. Larger holes and thinner septa increase the number of transmitted photons, so the sensitivity is higher. However, more oblique photons are accepted, which worsens spatial resolution. Smaller holes and thicker septa improve resolution, but many photons are stopped, so sensitivity is lower.
In simulation, you represent the collimator as a volume of high density material, usually lead. The hole pattern is then modeled explicitly, or through repeated unit cells. The collimator must be placed very close to the crystal and must cover at least the full active detector area. When you build a gamma camera geometry in GATE, it is important to align the collimator holes with the imaging direction, because the angular response of the collimator defines the mapping from activity in the patient to detected counts.
A crucial physical effect related to the collimator is septal penetration. Photons with sufficiently high energy, relative to the collimator design energy, can pass through the septa instead of being absorbed. This introduces unwanted counts from photons that did not travel along the accepted directions, which degrades image contrast and quantitative accuracy. In GATE, realistic modeling of the collimator material, thickness, and hole geometry allows you to study septal penetration, scatter in the collimator, and their impact on image quality.
The collimator defines the directional sensitivity of a SPECT scanner. Its geometry and material properties are critical for spatial resolution, sensitivity, and image contrast.
Scintillation crystal
Behind the collimator sits the scintillation crystal, which converts incoming gamma photons into visible light. In most conventional SPECT gamma cameras, the crystal is a large, flat slab of thallium doped sodium iodide, written as NaI(Tl). This material has a high light yield, relatively good energy resolution for gamma rays in the 100 to 300 keV range, and can be grown in large continuous plates.
When a gamma photon passes through the collimator and interacts in the crystal, it deposits energy. This energy deposition excites the crystal lattice and, through scintillation processes, leads to the emission of many optical photons. The total amount of scintillation light is approximately proportional to the energy deposited. For example, a 140 keV photon from Tc 99m interacting in NaI(Tl) will produce thousands of optical photons.
The scintillation crystal in a clinical gamma camera is usually several hundreds of millimeters in lateral dimensions and about 9 to 12.7 mm in thickness. The thickness is a trade off between detection efficiency and intrinsic spatial resolution. A thicker crystal absorbs more photons, so the detector is more efficient, but the broader depth distribution of interactions increases parallax effects and intrinsic blurring. In simulation, you choose the crystal thickness according to the desired energy and efficiency. For low energy SPECT, like Tc 99m at 140 keV, a standard thickness is often sufficient.
From a simulation perspective, you can treat the crystal at two different levels. In non optical simulations, you consider only the gamma interactions and the total energy deposited in the crystal volume. This is often enough for energy spectra, count rates, and basic imaging performance. In optical simulations, you also model the production and transport of optical photons within the crystal, including reflection at surfaces and absorption in the material, to study light sharing and photodetector response in detail.
The crystal is typically optically coupled to an array of photodetectors. These may be traditional photomultiplier tubes, or arrays of silicon photomultipliers in modern systems. The way light spreads within the crystal and reflects on its surfaces determines how the system estimates the interaction position from the pattern of light seen by multiple photodetectors. Although detailed light transport is part of optical photon simulation, it is still useful, even in simple models, to understand that the scintillation crystal is the main site of energy deposition and the origin of the electronic signal.
The scintillation crystal converts gamma ray energy into visible light. Its material, thickness, and size directly influence detection efficiency, energy resolution, and intrinsic spatial resolution.
Detector head
The detector head is the complete mechanical and functional assembly that contains the collimator, scintillation crystal, light guide if present, photodetectors, shielding, and associated housing. In clinical SPECT scanners, each detector head is a large, flat unit that rotates around the patient. Many systems have one, two, or three detector heads mounted on a gantry, allowing different angular configurations.
From the point of view of geometry in GATE, the detector head is a hierarchical structure. At the top level you have a detector head volume that defines the overall external dimensions and position. Inside this volume, you place the collimator at the front, the scintillation crystal directly behind it, and then the photodetector region, usually represented as one or more volumes that mimic photomultiplier tubes or solid state detectors. Behind and around these active components you add shielding volumes, typically lead, to reduce background from photons entering from outside the field of view.
The spatial relationship between these internal components matters. The collimator must be tightly coupled to the crystal with a small gap, since this gap influences the effective resolution and parallax. The distance from the patient to the collimator face, often called the detector to patient distance, is controlled by the scanner mechanics and has a direct effect on spatial resolution and sensitivity. In simulation, this distance is simply the gap between the patient or phantom volume and the detector head volume.
Many SPECT systems rotate the detector head around the patient to acquire projections at multiple angles. In a GATE simulation, you implement this by rotating the detector head volume relative to the world or patient volume over time. This motion is an important part of SPECT acquisition and is treated in more detail in later chapters. Here, you only need to keep in mind that the detector head is the object that moves, carrying the collimator and crystal together as a rigid assembly.
The detector head is also where you define the readout structure for data analysis. You may give the detector head an identifier, and optionally define sub volumes for finer segmentation, for example virtual pixels or modules. These identifiers are later used by actors and digitizers to map energy deposits and hits to specific parts of the detector, which is essential when you want to reconstruct projection images or study detector uniformity.
Finally, the detector head contains passive components that are important in realistic simulations, such as mechanical support plates and shielding blocks. These components contribute to scatter and attenuation before photons reach the crystal. If you neglect them, you may overestimate system sensitivity or underestimate object scatter. In practice, you balance geometric detail against simulation time. For basic educational simulations, a simple detector head with a collimator and crystal may be enough. For research on quantitative accuracy or system design, a more complete detector head model is recommended.
The detector head is the complete SPECT detection unit, including collimator, scintillation crystal, photodetectors, and shielding. Its geometry, position, and motion define how projection data are acquired in SPECT.
Views: 10