KAHIBARO
30.5. SPECT Source
Table of Contents
Point source
In SPECT simulations, the simplest activity distribution you can define is a point source. A point source represents all radioactive activity concentrated in a very small region of space, typically treated as a mathematical point compared with the size of the detector and phantom. In practice, the physical source may have finite dimensions, but if these dimensions are negligible for the imaging problem, a point model is appropriate.
In GATE, a point SPECT source is defined by specifying a fixed position, an isotropic or directional angular distribution, an energy spectrum that matches the radionuclide, and an activity that determines how many photons are emitted over time. For SPECT radionuclides such as Tc-99m, you typically model a monoenergetic primary gamma near 140 keV, or a discrete set of gamma lines with appropriate branching ratios. The detailed radionuclide physics is defined elsewhere, so for the source definition you usually choose either a monoenergetic gamma beam at the photopeak energy or a radionuclide decay source if you want to include all emitted photons.
A point source is especially useful for characterizing gamma cameras and SPECT systems. It allows you to study intrinsic detector properties without the complexity of extended activity distributions. Typical applications include measuring spatial resolution, energy resolution, sensitivity, and uniformity. For resolution studies, you place the point source at known positions in front of the detector head or within a simple water phantom and record the detected projection data. By reconstructing or by analyzing the projection profile directly, you can determine the system’s point spread function, usually characterized by the full width at half maximum of the response.
When you create a point source for SPECT, you must choose its location carefully relative to the collimator and detector. If the source is in air at a known distance from the collimator, it can help you measure planar system response. If you embed the source at specific depths inside a water or tissue-equivalent phantom, you can study attenuation and scatter effects separately from pure geometric blurring. For basic beginner simulations, placing the point source on the axis of the gamma camera and varying the source to collimator distance is a convenient way to see how spatial resolution and sensitivity change.
Another important choice is the angular distribution of emitted photons. For SPECT validation and simple camera characterization, an isotropic point source is usually used, which means the photons are emitted uniformly in all directions. Only the fraction within the solid angle of the collimator holes contributes to the detected signal. For more specialized tests you might restrict the angular distribution to a cone, but this is less common in standard SPECT camera studies.
In an OpenGATE Python script, a point source is defined by assigning the source type, particle type, position, and activity, as well as any relevant energy distribution. While the code details belong in other chapters, it is important to understand conceptually that the point source is just one source object with a fixed coordinate in the world volume. As you move the source or adjust its activity, you immediately affect the resulting projection data and count statistics in the simulation.
Point sources also play a role in checking that your SPECT geometry and physics are set up correctly. A misplaced point source or an incorrect distance to the detector can reveal problems in collimator definition, detector orientation, or coordinate systems. Because the expected result is simple and well understood, point source simulations are an excellent first test when building a new SPECT model.
For SPECT system characterization, always use a well positioned isotropic point gamma source with a correctly defined energy corresponding to the radionuclide, for example about 140 keV for Tc-99m, and verify the source to collimator distance carefully.
Phantom source
A phantom source represents a distributed activity within an extended volume, such as a uniform cylinder, an anthropomorphic phantom, or a patient-like object. Unlike the point source, which concentrates activity at a single location, a phantom source models how SPECT systems image realistic activity distributions that vary in space.
In SPECT simulation, phantom sources are essential when you want to study image quality, contrast, quantification, attenuation, and scatter in clinically relevant conditions. Typical examples include a uniform cylindrical phantom, a hot sphere in a cold background, or more complex anthropomorphic phantoms with organ structures and different activity levels. The same basic concept applies: you define a volume that represents the physical phantom and then specify how the activity is distributed inside that volume.
In GATE, a phantom source is usually implemented as a volume source. The emission positions are sampled randomly within the boundaries of a geometric object that you have already defined in the geometry section of your simulation. For a simple uniform phantom, you assign a constant activity concentration throughout the entire volume. During the simulation, individual gamma emission positions are drawn according to this uniform distribution, so the overall effect is that the whole volume emits photons at a constant rate per unit volume.
You can model more complex SPECT phantom sources by combining multiple volumes, each with its own activity, or by defining non uniform activity distributions. For instance, to simulate hot and cold regions, you might give one region a high activity concentration and another region a low or zero activity. This configuration can be used to evaluate contrast recovery, lesion detectability, and the influence of partial volume effects after reconstruction.
When you work with anthropomorphic phantoms or patient-like phantoms, you can assign different activities to different organs or regions. Each region behaves like its own volume source, and together they form a composite phantom source. The total activity is divided according to the specified activity concentrations and volumes, and the gamma emissions originate from the entire set of active regions. This allows you to reproduce realistic uptake patterns for SPECT tracers, although the detailed mapping from biological uptake to activity distribution is handled in other chapters.
A crucial choice for phantom sources is the activity concentration, which is typically given in units such as Bq per milliliter. Once you specify the activity concentration and the volume, you can compute the total activity in the phantom as
$$
A_\text{total} = C_\text{act} \times V_\text{phantom},
$$
where $C_\text{act}$ is the activity concentration and $V_\text{phantom}$ is the volume. The simulation then uses the total activity, together with the acquisition time, to determine how many gamma emissions should be generated.
For extended SPECT phantoms, define the activity as a concentration in Bq per milliliter, compute the total phantom activity with $A_\text{total} = C_\text{act} \times V_\text{phantom}$, and ensure that the volume used in this calculation exactly matches the volume used as the source region.
In SPECT simulations, phantom sources inherently include attenuation and scatter when they are placed inside attenuating materials such as water, soft tissue, or bone. By placing the phantom volume at a clinically realistic distance from the detector and surrounding it with appropriate materials, you can study how attenuation correction and scatter correction affect reconstructed images. For example, a uniform cylindrical water phantom with uniform activity is often used to study uniformity and quantification. Hot spheres inside such a phantom are used to evaluate contrast and recovery coefficients.
From a practical perspective in OpenGATE, setting up a phantom source involves two main steps. First, you define the phantom geometry and assign materials to its volumes. Second, you create one or more volume sources attached to those volumes, specifying the radionuclide or gamma energy, the activity or activity concentration, and the spatial distribution within the volume. In simple uniform cases, the position sampling is uniform throughout the volume. In more advanced cases, you can use image based distributions or masks, which are handled in other chapters.
Phantom sources are fundamental when you want to move from basic system testing with point sources to realistic performance evaluation. They allow you to evaluate not only the detector and collimator response, but also the combined effects of attenuation, scatter, and reconstruction algorithms on clinically relevant activity distributions.
Views: 11