KAHIBARO
5.1. Physical Units
Table of Contents
Length
In GATE, every quantity must be given together with an explicit unit. Length is used for all geometry descriptions, source positions, detector sizes, and scoring grids. GATE follows the Geant4 convention and provides a hierarchy of metric units based on the meter.
The base unit of length is the meter, written as m. From this, other units are defined as constants in Python, for example mm, cm, and km. When you define a size or a position, you always multiply by the unit constant, for instance 10 mm or 5 cm. This multiplication converts a human readable value into internal simulation units.
Typical length scales in medical physics simulations cover many orders of magnitude. For example, detector crystals and small phantoms are often a few millimeters, patient bodies are tens of centimeters, and room shielding can be meters thick. In GATE it is normal to mix these in one simulation, because the unit constants keep everything consistent.
In practice, this means that if you want to define a cubic detector crystal with a side of 4 mm, you write something like:
crystal.size = [4 * mm, 4 * mm, 20 * mm]
Use small length units such as mm or cm for human scale objects, and remember that positions use the same units as sizes. This keeps geometries easier to understand and reduces mistakes from confusion between different scales.
Always multiply numeric values by the appropriate length unit constant, such as mm, cm, or m. Never enter bare numbers for distances, positions, or sizes.
Energy
Energy units describe the energy of particles, especially photons, electrons, positrons, and protons, and also appear in thresholds and spectra. The base unit of energy in GATE and Geant4 is the electron volt, written as eV.
Medical physics simulations commonly use multiples of the electron volt. For nuclear medicine and PET, energies are typically in the range of hundreds of keV, such as 140 keV for Tc 99m or 511 keV for annihilation photons. For radiotherapy and CT, energies are usually given in MeV or given as a kVp tube potential. These are expressed through constants such as keV and MeV in GATE.
When you assign an energy to a source, or define an energy window or a cut, you always use a numeric value multiplied by an energy unit constant. For example, to define a monoenergetic gamma source at 140 keV you might write:
source.energy.mono = 140 * keVor to specify an energy window from 400 keV to 650 keV, use:
digitizer.energy_window.min = 400 * keV
digitizer.energy_window.max = 650 * keVThis convention also applies to physics settings such as production cuts and step limits if they are defined in energy.
Always attach an energy unit, such as keV or MeV, to any energy value. Do not confuse energy in keV or MeV with dose in Gray, they are different quantities.
Time
Time appears in GATE in several ways, including acquisition duration, time stamps of events, timing resolution, and radioactive decay. The base unit of time is the second, written as s.
For practical simulations, shorter units are used very often, particularly for detector timing and time of flight PET. GATE provides constants such as ns for nanoseconds, us for microseconds, and ms for milliseconds. Time steps and acquisition intervals are then set by multiplying with these constants.
For instance, if you want to run a simulation that represents 10 seconds of acquisition, you can specify:
simulation.run_time = 10 * sIf you want to define a coincidence time window of 4 ns, you might write:
digitizer.coincidence.time_window = 4 * nsTime is also essential for time dependent activity and moving geometries. In those contexts, you use the same units for defining when a motion starts, how long it lasts, or how activity decays over the course of the simulation.
Use explicit time units such as ns, ms, or s for all durations and time windows. Mixing plain numbers with numbers that include units can silently create incorrect acquisition times and timing cuts.
Activity
Activity describes how many nuclear decays occur per unit time in a radioactive source. In GATE, activity is most often given in Becquerels, the SI unit, which is defined as decays per second. The base unit is Bq.
In medical imaging, activities are rarely on the order of 1 Bq, so multiples are used. Common units are kBq for kilobecquerel, MBq for megabecquerel, and in some contexts even GBq. GATE provides unit constants that correspond to these scales.
When configuring a source that uses activity instead of a fixed number of primary particles, you set the activity multiplied by the appropriate unit. For example, to model a PET source with an activity of 5 MBq you might write:
source.activity = 5 * MBqIf your simulation time corresponds to a real acquisition duration, the product of activity and time determines the total number of decays that can be generated statistically. GATE internally uses the activity and duration to sample decay times and to decide how many particles to simulate.
Activity units are tied to real world quantities. This means that if you want your simulation to be comparable to an experiment, you must ensure that both the activity and the time duration are correctly specified with their units, and that you understand how GATE interprets these values for radioactive sources.
Activity is measured in decays per second. Always specify activity with Bq, kBq, or MBq. Do not confuse activity with the number of simulated primary particles, they are related but not identical concepts.
Dose
Dose is a measure of energy deposited per unit mass, used to quantify the effect of radiation on matter, particularly in patient and phantom simulations. In SI units, absorbed dose is measured in Gray, written as Gy, which is defined as 1 Joule per kilogram.
In GATE, dose is not something you directly assign as an input to the simulation. Instead, you configure actors that score energy deposition and convert it to dose using the known densities and masses of the materials. These actors typically output dose maps in units of Gray or submultiples such as mGy.
The relationship between energy deposition and dose is:
$$
D = \frac{E_{\text{dep}}}{m}
$$
where $D$ is the dose in Gray, $E_{\text{dep}}$ is the energy deposited in Joules, and $m$ is the mass in kilograms. In practice, GATE handles this conversion internally within dose actors, starting from particle energies given in eV and using standard conversions between eV and Joule.
Although you do not usually multiply by Gy inside the simulation script, you must be aware of the units when you interpret output. Dose images exported by GATE will specify whether values are in Gy or in a scaled unit. When you analyze dose numerically, for example when extracting dose profiles or computing mean organ dose, you must keep track of these units to avoid errors by factors of 10 or 1000.
Dose is energy per unit mass and is reported in Gray. Treat dose results as physical quantities with units, and always check whether your output files use Gy, mGy, or another scale before comparing with measurements or constraints.
Angle
Angles in GATE describe orientations and rotations of objects, as well as directions of particles and detector motions. The fundamental unit for angles in Geant4 is the radian, although degrees are also commonly used through provided constants.
A radian is defined through the ratio of arc length to radius, and there are $2\pi$ radians in a full circle. In many configurations, it is convenient to think in degrees, for example when specifying that a gamma camera rotates from 0 to 360 degrees around the patient, or when defining collimator tilt angles. GATE therefore provides constants such as deg or similar degree related units that you multiply by a numeric value.
When you define a rotation, you often specify an angle and an axis. For instance, to rotate a detector head by 90 degrees around the z axis, you might set a rotation that internally uses $90 * deg$. Similarly, for SPECT or CT simulations, projection angles are often generated as a sequence of equally spaced degree values converted into radians by the constants.
Angles are not limited to macroscopic rotations. They also appear in definitions of particle directions and angular distributions, for example when simulating a cone beam or a collimated source. In these cases, you may define opening angles or polar and azimuthal angles in radians, again usually through unit constants.
Angles are handled in radians internally, but you can use degree constants for convenience. Always attach an angle unit, such as deg or rad, when specifying rotations or angular ranges to avoid incorrect orientations.
Views: 11