Abstract
Proton therapy is gaining popularity in the world. Proton beam possesses the physical property of Bragg peak that can be used to provide higher radiation dose to tumor and spares the normal tissues. However, the inaccuracy in proton range limits the advantages of Bragg peak. The proton range verification plays a very important role in the proton therapy. In recent years, camera equipped with collimators (multi-slit and knife-edge slit collimator) to image prompt gamma (PG) emitted along the proton tracks in the patient have been proposed for range verification. The aim of the work is to develop a system which combines a gamma camera for tumor localization and collimator-based camera for proton range verification. This study can be separated into three stages. GATE/GEANT4 will be used as a simulation tool of the proton therapy in all stages. The purpose in the first stage is to compare the performance of multi-slit collimator and knife-edge slit collimator in range verification. Factors including the energy window setting, proton energy, phantom size, and phantom shift that may influence the accuracy of detecting range were studied. In the second stage, we evaluate the feasibility of gamma camera for tumor localization in proton therapy. The tumor size, phantom shift and the depth of tumor localization will be studied in a rectangular PMMA phantom. Finally, in the third stage, we combine gamma camera with the selected range verification system (from stage one) to evaluate this method based on a simulated tumor embedded in a Zubal phantom. Results in the first stage indicate that both collimator systems have good response to the phantom shift, and knife-edge collimator system achieve higher detection efficiency leading to a smaller deviation in predicting range. In the stage two, gamma camera achieves better accuracy of detecting the phantom shift than knife-edge system. The final stage indicate this method is feasible with SPECT reference image. However, the quality of gamma camera image should be improved by other scanning modes. We conclude that combining gamma camera and collimator-based camera have potentials for accurately range monitoring in proton therapy. It is noted that neutron contamination has a marked impact on range prediction of the collimator-based camera, especially in multi-slit system. Therefore, a neutron reduction technique for improving the accuracy of range verification of proton therapy is needed.