BL4S (Beam Line for Schools) 2026 — TED Ege College, Aydın, Turkey
Characterisation of Compton Scattering Geometry and Its Contribution to Timing Resolution in Time-of-Flight Detection Systems
This repository contains a Geant4 simulation of the experimental beamline proposed for CERN's Beam Line for Schools 2026 competition. The simulation models a sequential detection chain designed to quantify how the geometric variability of Compton-scattered electron trajectories contributes to the integral timing resolution of a time-of-flight (TOF) system — a question directly relevant to the design of next-generation TOF-PET scanners used in oncological imaging.
The central research question is:
To what extent do scattering geometry and path-length variation contribute to the integral timing resolution of the system, independently of intrinsic detector jitter?
The simulated geometry follows the proposed experimental setup in order of particle traversal:
| Position (z) | Component | Material | Purpose |
|---|---|---|---|
| −120 cm | Electron gun | — | 0.5 GeV e⁻ source |
| −90 cm | Cherenkov detector | CO₂ gas | Beam tagging & energy verification |
| −75 cm | Bremsstrahlung target | Tungsten (2.5 X₀ ≈ 8.75 mm) | Gamma photon production |
| −55 cm | Tissue phantom | ICRP soft tissue | Clinical reference medium |
| −35 cm | Collimator | Lead (5 cm, 1 cm aperture) | Beam definition |
| −20 cm | Scintillator 1 (t₁) | Polystyrene | First TOF timestamp |
| +5 cm | DWC1 | Argon gas | Track reconstruction |
| +30 cm | DWC2 | Argon gas | Track reconstruction |
| +60 cm | Scintillator 2 (t₂) | Polystyrene | Second TOF timestamp |
The tungsten target thickness of 2.5 X₀ was chosen to maximise Bremsstrahlung photon yield while minimising photon self-absorption via the photoelectric effect and pair production.
The simulation uses the QBBC physics list, which includes:
- Full electromagnetic physics (Livermore model for low energies)
- Hadronic interaction models (FTFP + Bertini cascade)
- Compton scattering via the Klein–Nishina model
- Bremsstrahlung with LPM effect above 1 GeV
The Compton kinematic relation governing the experiment is:
Δλ = λ' − λ = (h / mₑc)(1 − cosθ)
Each simulation run produces six CSV files:
| File | Contents |
|---|---|
volume_entries.csv |
Every particle entering each detector volume (eventID, particle, energy, position, process, origin volume) |
volume_edep.csv |
Energy depositions per volume per step |
detector_spectrum.csv |
Particle spectra at each detector |
photon_spectrum.csv |
Gamma-only subset of detector spectrum |
phantom_dose_distribution.csv |
3D dose map within the tissue phantom |
angular_distribution.csv |
θ and φ of gamma photons entering each volume |
These outputs are designed to allow the geometric timing contribution (σ²_geo) and the intrinsic detector jitter contribution (σ²_det) to be separated via:
σ²_total = σ²_geo + σ²_det
- Geant4 ≥ 11.0 (built with Qt and OpenGL for visualisation)
- CMake ≥ 3.16
- A C++17-compatible compiler
Ensure the following Geant4 datasets are installed: G4LEDATA, G4LEVELGAMMADATA, G4NEUTRONXSDATA, G4SAIDXSDATA, G4ENSDFSTATEDATA.
# Clone the repository
git clone https://github.com/<your-org>/<repo-name>.git
cd <repo-name>
# Create a build directory
mkdir build && cd build
# Configure
cmake ..
# Build
make -j$(nproc)./exampleB1The visualiser opens automatically. Trajectories are colour-coded: red = e⁻, green = γ, blue = e⁺.
# Short test run (5 gamma + 1 proton events, verbose tracking)
./exampleB1 run1.mac
# Full statistics run (10,000 events per particle type)
./exampleB1 run2.mac
# Predefined test with output redirect
./exampleB1 exampleB1.in > exampleB1.out.
├── include/
│ ├── ActionInitialization.hh
│ ├── DetectorConstruction.hh
│ ├── EventAction.hh
│ ├── PrimaryGeneratorAction.hh
│ ├── RunAction.hh
│ └── SteppingAction.hh
├── src/
│ ├── ActionInitialization.cc
│ ├── DetectorConstruction.cc
│ ├── EventAction.cc
│ ├── PrimaryGeneratorAction.cc
│ ├── RunAction.cc
│ └── SteppingAction.cc
├── exampleB1.cc # Main entry point
├── CMakeLists.txt
├── run1.mac # Interactive test macro
├── run2.mac # High-statistics batch macro
├── exampleB1.in # Predefined batch input
├── vis.mac # Visualisation settings
└── README.md
Compton scattering accounts for 70–80% of photon interactions in the energy range used in radiotherapy. In TOF-PET systems, the precision with which an annihilation vertex can be localised is bounded by timing resolution, expressed as:
Δx = c · Δt / 2
Improved timing resolution directly improves the signal-to-noise ratio and enables more targeted dose delivery to tumour tissue. This simulation provides a Monte Carlo baseline for separating the geometric path-length contribution from detector jitter — data that can inform the optimisation of sub-100 ps TOF-PET detectors.
TED Ege College, Aydın, Turkey — BL4S 2026
Arhan Hasan Ünsal · Atakan Korkmaz · Beren Duygu Yılmaz · Doruk Turan · Doruk Utku Tarim · Leman Ece Genclesen · Neva Yildizli · Kayra Sari · Arda Genc · Duru Sefa
Supervisor: Eda Erdogan
- Surti S, Karp JS. Advances in time-of-flight PET. Phys Med Biol. 2016.
- Joseph S, Matthijs V, et al. Bremsstrahlung, Synchrotron Radiation, and Compton Scattering. Rev. Mod. Phys. 1970; 42:237.
- Surti S, et al. Impact of TOF on PET/CT imaging. J Nucl Med. 2010; 51(2):237–245.
- Conti M, et al. Impact of TOF on PET tumor detection. J Nucl Med. 2009; 50(8):1315–1323.
- Tajima M, et al. Carbon target in 90° Compton spectroscopy. J Nucl Sci Technol. 2008; 45(8):760–765.
- Llosá G, et al. Noise evaluation of Compton camera imaging. Phys Med Biol. 2015; 60(5):1845.
- Shimazoe K, et al. TOF-PET using Compton scattering by plastic scintillators. VCI Proceedings. 2016.