26.4. PET Source

Table of Contents

F-18 source

In PET simulations with GATE, the PET source is usually a positron emitter placed in a phantom or patient model. For an introductory PET setup, the standard choice is Fluorine 18, written in GATE as F18 or F-18 depending on context. The radioactive decay and emission of positrons are handled by the physics configuration described in the PET Physics Configuration chapter, so here you focus on how to define the source itself.

In a typical OpenGATE Python script, you first create the simulation, then add a source and select the radionuclide. A minimal pattern for an F-18 PET source looks like this, assuming you already created sim and a world or phantom volume:

python

import opengate as gate
import opengate_core as g4  # name may differ, check your installation
from opengate.contrib import gate_actor
sim = gate.Simulation()
# ... geometry and physics configuration defined elsewhere ...
# create an F-18 source
src = sim.add_source("my_f18_source", "Generic")
src.particle = "e+"          # emitted positron; radionuclide decay handles creation
src.nucleus = "F18"          # tell GATE which radionuclide to use
src.activity = 1e6 * gate.Bq # 1 MBq, see Units in GATE chapter

The key idea is that you are not directly creating 511 keV photons. Instead, you tell GATE to simulate the decay of F-18 nuclei, which produces positrons. The positron slows down, may travel some distance in tissue, then annihilates with an electron and generates two 511 keV photons. This chain is essential for realistic PET, since it introduces positron range and non collinearity effects when combined with the PET Physics Configuration.

There are two common ways to specify a PET source in GATE. For educational or debugging purposes, you may sometimes bypass full radioactive decay and create a pure positron or gamma source with fixed energy. This is not a genuine PET radionuclide, but can be useful to check geometry or timing:

python

# simplified positron-only source, not a real radionuclide decay
src = sim.add_source("test_pos", "Generic")
src.particle = "e+"
src.energy.mono = 0.64 * gate.MeV   # typical F-18 endpoint energy

For realistic PET simulations, you should instead use the radioactive source approach. In OpenGATE, this is usually done either by specifying src.nucleus as shown above, or by selecting a specific radioactive source type if your version provides a radionuclide source helper. The exact field names can vary slightly between OpenGATE releases, so always check a recent example from the GATE documentation or bundled PET examples.

Another important aspect is the spatial and angular definition of the F-18 source. Even for a simple PET scanner test, you must place the activity inside a volume, for example a cylindrical phantom in the center of the scanner. The position and direction settings of the source control where decays occur and how emitted particles are oriented. In PET, decays are isotropic, so you generally choose an isotropic or default angular distribution, not a fixed beamlike direction. A minimal centered point source looks like this:

python

src.position.type = "point"
src.position.center = [0, 0, 0]       # center of the world or phantom
src.direction.type = "iso"            # isotropic emission

The actual F-18 decay scheme and branching ratios are handled by the Geant4 radioactive decay models you enable in the PET Physics Configuration, so you do not need to manually define the decay gammas or beta spectrum. Your task in this chapter is mainly to create the PET source object, select F-18 as the radionuclide, and connect it to the simulation geometry through position and activity settings.

For realistic PET simulations:

Use a radioactive F-18 source, not a simple gamma source at 511 keV.
Place the activity inside a physical volume that is already defined in your geometry.
Use isotropic emission for the decay, not a fixed beam direction.

Once you can generate a basic F-18 source in the center of your PET ring, you can extend this to more complex distributions, such as line sources, phantom inserts, or patient specific activity maps, which are covered in the next section.

Activity distribution

In PET, the activity distribution describes how the decay rate of the radionuclide varies in space. Even if you use the same radionuclide F-18, the resulting images and detector data depend strongly on whether the activity is concentrated in a small sphere, spread uniformly in a cylinder, or follows a complex pattern from a patient image.

At the simplest level, you control the activity distribution by choosing the source position type and its spatial parameters. In OpenGATE, a point source has all activity at a single location, a box or sphere distributes activity uniformly in that volume, and a volume based source follows the shape of an existing geometry volume.

For example, a uniform F-18 distribution inside a cylindrical water phantom that you previously defined could be created as follows:

python

src = sim.add_source("f18_phantom", "Generic")
src.particle = "e+"
src.nucleus = "F18"
src.activity = 5e6 * gate.Bq    # 5 MBq total
src.position.type = "volume"
src.position.translation = [0, 0, 0]
src.position.confine = "water_phantom"   # name of the phantom volume

Here, the total activity is 5 MBq, distributed uniformly over the specified volume. Each decay position is sampled randomly within the phantom. The PET image you reconstruct later will reflect this uniform activity, filtered by attenuation and scanner response.

For testing scanner resolution, you may instead choose a small spherical hot spot. In that case, you can use a sphere position type or define a small spherical volume in the geometry and confine the source to it. A simple spherical distribution without defining a separate volume can look like:

python

src.position.type = "sphere"
src.position.center = [0, 0, 0]
src.position.radius = 5 * gate.mm      # 10 mm diameter hot sphere

This creates a small region of activity ideal for evaluating spatial resolution or point spread.

For more realistic studies, especially when working with anthropomorphic phantoms or patient data, activity distributions become voxel based. Instead of a simple analytical shape, you load an image where each voxel contains an activity value or a relative uptake. GATE can then sample decay positions according to this voxelized map. The technical details of voxelized activity maps and image based sources are covered in the Internal Dosimetry and Voxelized Geometry chapters, but the PET perspective is that such maps represent tracer uptake patterns in tissues or tumors.

You should also be aware of the relationship between activity distribution and acquisition time. Activity is given in becquerel, where 1 Bq is 1 decay per second. If you define a total activity $A$ and simulate for an acquisition duration $T$, the expected number of decays is approximately:

$$
N_{\text{decays}} \approx A \times T
$$

In GATE, you can either specify the total number of simulated events directly or define an activity and a simulation time. The PET Simulation Fundamentals chapter on PET Physics Configuration discusses how to connect activity, time, and number of events. For the source definition itself, it is important that your chosen activity distribution and acquisition settings result in a reasonable number of events for both physics accuracy and computation time.

Important rules for PET activity distributions:

Uniform shapes such as points, spheres, and boxes are ideal for scanner tests and validation.
Use volume confinement when you want the activity to follow an existing phantom or insert.
Voxelized activity maps are required for realistic tracer uptake in patient or organ phantoms.
The total number of decays is controlled by both activity and acquisition time, not by activity alone.

By combining an F-18 radionuclide definition with an appropriate activity distribution, you can construct PET sources that match basic test phantoms, standardized NEMA studies, or realistic clinical uptake patterns. Subsequent chapters on PET Detector Response and PET Coincidence Analysis will show how these source choices influence measured singles, coincidences, and reconstructed image quality.

26.4. PET Source

F-18 source

Activity distribution

Comments

Where to Move