KAHIBARO
34.1. External Beam Radiotherapy
Table of Contents
Treatment beams
External beam radiotherapy uses beams of radiation generated outside the patient and directed into the body to treat tumours. In GATE you model these treatment beams as particle sources with specific geometry, energy, and time characteristics that mimic clinical linac beams or other external sources.
A treatment beam for photon radiotherapy is usually represented as a rectangular field defined at some distance from the source, typically the isocenter. In a Monte Carlo simulation you do not need to reproduce every technical detail of the accelerator head in this chapter, that belongs to more advanced beam modeling, but you do need to capture the key physical features of the beam that affect dose in the patient.
The first essential property is the particle type. For megavoltage photon treatments you will typically choose primary photons with energies in the MeV range, such as 6 MV or 10 MV nominal beams. In GATE you describe this by specifying an energy distribution, not a single monoenergetic value, because clinical beams have a spectrum of photon energies. A simple starter model might use a monoenergetic beam to illustrate the concepts, but for realistic dose calculations you should use an energy spectrum that has been either measured, taken from the literature, or generated by a separate accelerator head simulation and stored as a spectrum or phase space.
Next you define the beam direction and shape. Most linac beams used in external beam therapy are approximately parallel over the field area at the patient, since the source is far compared with the field size. In GATE you typically implement a source located at the linac target position, pointing toward the isocenter. The field shape can be approximated using a rectangular or square source distribution in the plane perpendicular to the beam axis to represent a collimated field. The field size is a clinical parameter, such as 10 cm by 10 cm at isocenter, and is directly related to the lateral dose profiles you will score in the phantom or patient geometry.
Beam divergence is an inherent property of a point source located at a finite distance. If you position the source at a finite distance upstream from the isocenter, GATE will automatically produce a diverging beam. The apparent field size changes with depth and off‑axis position according to the geometry you define. For more advanced configurations, you might reproduce jaw or multileaf collimator openings by defining complex apertures, but for beginners a simple rectangular distribution is sufficient to understand how geometrical field size and divergence influence the dose.
The number of particles or beam intensity is related to the delivered monitor units in the clinic. In Monte Carlo you usually specify either the number of primary histories or an equivalent time or activity that controls how many particles are simulated. For external beam radiotherapy the absolute dose per monitor unit can be obtained by normalizing your simulated dose to a reference measurement, providing a link between the Monte Carlo output and clinically used dose prescriptions.
Treatment beams can also be shaped and modulated. Modern techniques such as intensity modulated radiotherapy or volumetric modulated arc therapy rely on spatial and temporal modulation of the beam intensity and field shape. In GATE this is represented by either dynamic changes in source parameters or by using multiple sources or configurations that correspond to different control points or segments. For an introductory treatment beam model you will typically start with a static, non‑modulated field to understand fundamental relationships such as depth‑dose curves and penumbra formation.
Finally, beam quality is not only determined by the primary photon spectrum but also by scattered radiation produced in the accelerator head. In basic beam models where the accelerator head is not explicitly simulated, you can approximate this head scatter by slightly broadening the energy and angular distributions, or by using spectra and angular distributions derived from more complete simulations. As you develop more advanced models, you might replace the simplified source with a phase space file produced by a detailed head simulation, and then use this phase space as a source in patient or phantom simulations.
For external beam radiotherapy dose calculations, the treatment beam model must correctly represent:
- The photon energy spectrum, not just a single energy.
- The field size at isocenter and beam divergence.
- The beam direction and source‑to‑isocenter distance.
Incorrect beam energy or geometry leads to wrong depth‑dose curves and lateral profiles, even if the rest of the simulation is correct.
Patient geometry
In external beam radiotherapy simulations the patient geometry describes how radiation interacts with the patient’s body and determines the spatial distribution of absorbed dose. In GATE this geometry can range from a simple homogeneous water phantom to a full patient model based on CT images, with realistic anatomical structures and material mapping. The level of detail you choose should match the goals of your simulation and your available input data.
The simplest patient representation is a rectangular water phantom that approximates soft tissue. This is useful for learning how to configure the treatment beam and dose scoring, and for comparing your simulated depth‑dose and profile curves against reference data such as commissioning measurements. You define the phantom as a volume placed in the world, define its size to cover the field and depth of interest, and assign a material such as liquid water. By placing the beam so that it enters one face of the phantom, you can compute basic beam characteristics that are essential in radiotherapy, such as percentage depth dose and lateral dose profiles at different depths.
To move closer to a realistic patient, you can replace the simple phantom with an image based geometry derived from a CT scan. Instead of a single material, each voxel in the CT is mapped to a material and density according to its Hounsfield unit. This voxelized geometry approach allows you to account for differences between bone, lung, and soft tissue and to capture the actual shape and location of the tumour and organs at risk. The simulation then computes dose in a three dimensional grid that corresponds to the CT voxels, which can later be compared to treatment planning system dose distributions.
When you set up patient geometry for external beam radiotherapy, the spatial relationship between the patient and the beam is crucial. In the clinic, treatments are usually delivered with the patient positioned so that the target is placed at the linac isocenter. In your simulation you must ensure that the patient geometry is correctly positioned with respect to the coordinate system and the beam direction. This includes alignment of the CT coordinate system, if you use one, with the simulation world coordinates, and matching the isocenter position used in the treatment plan with the point you use for the beam in GATE.
Patient geometry may also include support structures such as the treatment couch, immobilization devices, or other accessories if they significantly affect the dose distribution. These can be modeled as additional volumes around or under the patient, with appropriate materials such as carbon fiber for the couch top. For many educational examples you may choose to ignore these structures and focus on the main patient volume to keep the geometry simple.
In external beam simulations you typically attach dose scoring actors to the patient geometry. These actors record absorbed dose in voxels covering the volume of interest, for example the entire patient or a smaller region around the target. The geometry definition therefore needs to be compatible with the dose grid you choose. In a homogeneous phantom the phantom itself might define the scoring grid, while in a CT based geometry the voxel grid from the image usually defines the scoring resolution and extent.
If you want to represent the tumour and organs at risk explicitly, you can create additional regions within the patient geometry. In a simple phantom these regions might be embedded volumes with different materials or labels that mark target and critical structures. In an image based model they may correspond to contours or masks derived from structure sets. Although contour handling belongs to a dedicated chapter, it is important to understand here that your external beam radiotherapy simulation can assign different analysis regions inside the same geometry to report mean dose or dose volume histograms for specific organs.
Patient geometry is inherently three dimensional and may be non uniform in density and composition. This complexity is exactly why Monte Carlo simulation is powerful in radiotherapy, since it can accommodate arbitrary geometries and material distributions without approximation. As you progress from simple phantoms to realistic patient models, you increase the fidelity of the dose calculation, but you must also be more careful with image orientation, coordinate transformations, and the consistency between the simulated beam setup and the clinical plan you wish to emulate.
For external beam radiotherapy, correct patient geometry requires:
- An appropriate representation of anatomy, from simple water phantoms to CT based voxelized patients.
- Accurate positioning of the patient relative to the beam isocenter and coordinate system.
- Consistent material and density assignment within the geometry.
Even a small misalignment between patient geometry and beam can lead to clinically significant dose errors in the target and nearby organs.
Views: 9