From Cyclotrons to Cancer Therapy: A Physics-Based Simulation of Proton Beam Interactions in Human Tissue | RISE Research

This paper reviews the physics underlying cyclotron-based proton beam therapy for cancer treatment and presents an original, computationally lightweight 3D Python simulation modeling proton interactions with human tissue. The literature review traces the cyclotron's development from Ernest Lawrence's 1929 concept through relativistic-effect challenges (which necessitated synchrocyclotrons, isochronous cyclotrons, and superconducting cyclotrons) to its modern medical applications, focusing on why the proton's Bragg peak property, a sharp dose peak at a targetable depth followed by minimal exit dose, makes proton therapy more tissue-sparing than conventional radiation. The paper identifies range uncertainty (unpredictability in exactly where protons stop in tissue) and high capital/operational costs ($100-200 million for typical multi-room centers) as the main barriers to wider proton therapy adoption, particularly in low- and middle-income countries. To address a gap in accessible modeling tools, the author built a 3D Monte Carlo-style simulation that models tumor and healthy cell kill outcomes based on beam energy, beam energy spread, angular spread, breathing-induced tumor motion, tissue density heterogeneity, and bone obstruction, accounting for electronic stopping power and both elastic and inelastic nuclear collisions. The simulation supports both single-session and fractionated treatment modeling and includes a fractionation-schedule optimizer, which identified a 24-fraction, 1.6 Gy-per-fraction schedule (with specific beam energy, spread, and spot-size parameters) that killed 99.1% of tumor cells while minimizing healthy cell damage. Unlike heavier physics-simulation platforms like GEANT4, the tool is code-free and designed for educational and exploratory rather than clinical use, helping users visualize how specific beam parameters affect range uncertainty and tumor coverage. The model and code are publicly available on GitHub and archived on Zenodo.