KAHIBARO
SPECT Physics
Table of Contents
Gamma-emitting radionuclides
Single Photon Emission Computed Tomography, or SPECT, is based on radionuclides that emit individual gamma photons. In contrast to PET, which relies on pairs of 511 keV photons from positron annihilation, SPECT uses radionuclides that undergo nuclear transitions and emit characteristic gamma rays at specific energies.
Many medical radionuclides used in SPECT are produced in nuclear reactors or cyclotrons and are chemically attached to tracers that follow specific biological pathways. Once inside the body, these radionuclides decay and emit gamma photons that escape the patient and are detected by the scanner.
The basic nuclear process is usually an isomeric transition or a beta decay that leaves the daughter nucleus in an excited state. The excited nucleus then de-excites and emits one or more gamma photons. Each radionuclide has its own set of discrete photon energies and branching ratios. For example, Tc-99m emits a dominant photon around 140 keV, which is widely used because it offers a good compromise between patient dose and image quality.
In SPECT simulations with GATE, you represent these gamma emissions in two conceptually different ways. You can use explicit nuclear decay physics where the code samples from a full decay scheme, or you can define a simplified monoenergetic gamma source at the relevant photon energy. The first approach is closer to reality and can include multiple gamma lines and possible emissions from daughter radionuclides. The second approach is often sufficient for scanner design and basic imaging studies and is computationally simpler.
An important concept is the gamma emission spectrum. Real radionuclides may emit several gamma lines, each with a probability given by its branching ratio. For example, some SPECT radionuclides emit photons at 140 keV, 159 keV, or higher energies, with different probabilities. In a simulation, this is represented as an energy distribution. Depending on the study, you might include all relevant lines or only the main photopeak.
Gamma photons interact with matter mainly through the photoelectric effect and Compton scattering in the SPECT energy range, typically between about 70 keV and 300 keV. The probability for each interaction type depends strongly on photon energy and material composition. This has direct consequences for both patient dose and image quality. In GATE, the physics lists you choose for SPECT must correctly model low energy photon interactions in tissues, collimators, and detector crystals.
In clinical practice, gamma-emitting radionuclides are chosen to combine suitable physical half-life, appropriate photon energy, and useful biological behavior. The physical half-life determines how long the tracer can be imaged and influences counting statistics. In simulation, the half-life also matters if you model time-dependent activity or dynamic acquisitions, although for short scans and static phantoms you often assume constant activity.
Because different radionuclides have different photon energies, they require different collimator designs and detector settings. A low-energy radionuclide like Tc-99m uses low-energy high-resolution collimators, while higher energy photons from radionuclides such as I-131 require thicker septa and different collimator geometries to limit septal penetration. When you simulate SPECT physics, the choice of gamma energy must be consistent with collimator material, thickness, and detector response.
In summary, SPECT physics begins with understanding that you are dealing with single gamma photons from specific nuclear transitions. Their energy spectrum, branching ratios, and time evolution define the photon field that leaves the patient and interacts with the SPECT camera. GATE then tracks these photons through collimators, shielding, and crystals to predict how many are detected and where.
In SPECT simulations, always ensure that the gamma energy spectrum of your source matches the radionuclide of interest, including the dominant photopeak energy and significant additional lines. Incorrect energies or missing lines can produce unrealistic scatter fractions, count rates, and image contrast.
Gamma camera imaging
A SPECT system uses a gamma camera to convert escaping photons into measurable electronic signals and finally into images. Understanding the basic physics of gamma camera imaging is essential before you configure geometries, sources, and physics lists in GATE.
A classical gamma camera consists of three main physical components in series: a collimator, a scintillation crystal coupled to light guides and photodetectors, and surrounding shielding. Gamma photons first encounter the collimator, which is usually made of a high atomic number material such as lead. The collimator contains many narrow holes separated by absorbing septa. Only photons that travel along preferred directions can pass through the holes and reach the crystal. All other photons are ideally absorbed in the septa.
The collimator is the key element that provides spatial information in SPECT. There is no focusing lens as in optical imaging. Spatial resolution is obtained by geometric selection of photon directions. This introduces a fundamental trade-off between sensitivity and resolution. Narrow long holes with thick septa improve angular and spatial resolution but reject more photons, reducing sensitivity. Wider or shorter holes increase sensitivity but degrade resolution and allow more scattered or misdirected photons to reach the detector.
Once a photon passes through the collimator it enters the scintillation crystal, typically NaI(Tl) in conventional gamma cameras. In the crystal, the photon may be absorbed via the photoelectric effect or undergo one or more Compton scatter events before being absorbed. The deposited energy is converted into a burst of scintillation light. The total light output is approximately proportional to the deposited energy, but with an intrinsic energy resolution that depends on crystal and detector properties.
The scintillation light is then collected by an array of photomultiplier tubes or other light sensors on the back of the crystal. By comparing light signals in different sensors, electronics estimate the interaction position within the crystal and the total energy deposited. The simplest position estimate uses weighted sums of photodetector signals, and more advanced methods perform more sophisticated position reconstruction. Timing information is less critical in standard SPECT than in PET, but acquisition systems still record event times to some extent.
Gated energy selection plays a central role in SPECT physics. The detector does not accept all events. It applies an energy window around the photopeak of the radionuclide. For example, for Tc-99m you might accept events within a certain percentage around 140 keV. Events with too low energy are likely due to Compton scatter in the patient, collimator, or detector, which have lost part of their energy and would cause image degradation. Events with too high energy may be noise or pile-up. In simulation, you often implement this energy selection via digitizers that blur energies and apply thresholds and windows.
The creation of a SPECT projection image is fundamentally a counting process. For a given camera orientation, the system records the number of accepted events in each detector region or pixel. Each count corresponds to at least one gamma interaction that has passed the collimator and satisfied the energy window. Because of the collimator geometry, each pixel is sensitive mostly to activity within a particular projection line through the patient. By rotating the gamma camera around the patient, you acquire many projection images from different angles. These are later reconstructed into a three dimensional activity distribution.
In GATE simulations of SPECT physics, you usually model photon transport from the source distribution through the patient or phantom, then through the collimator and crystal. The physics of photon attenuation and scatter in tissues, collimator absorption and septal penetration, and energy deposition in the crystal all influence the final detected spectrum and spatial distribution. Accurate low energy electromagnetic physics and realistic material definitions are therefore important.
One of the most subtle SPECT physics effects is scatter. Many photons scatter in the patient or in the collimator and still enter the detector, often with reduced energy and misdirected path. Even with energy windows, some scattered photons are accepted and contribute to image background and reduced contrast. Simulations help quantify the scatter fraction and design energy windows, collimators, and reconstruction corrections.
Another critical effect is septal penetration and collimator scatter, where photons pass through septa or scatter around them instead of going through the intended hole. This is more important for higher energy photons or for thin septa. It creates blur and background in the image. Simulating detailed collimator geometry in GATE lets you evaluate how collimator design and radionuclide energy interact to affect image quality.
Finally, detector intrinsic resolution and energy resolution define how precisely you can localize events and discriminate scattered photons. In GATE, these are usually modeled with digitizers that apply spatial and energy blurring to the underlying energy deposition information. This separates the pure interaction physics from the electronics and reconstruction aspects of gamma camera imaging.
A realistic SPECT simulation must include three key physics components: photon attenuation and scatter in the patient, collimator interaction including septal penetration, and energy deposition in the scintillation crystal with appropriate energy resolution and selection windows. Omitting any of these can significantly misrepresent system sensitivity, spatial resolution, and scatter fraction.
Views: 11