A comparison of Monte Carlo simulations of dual-energy and proton computed tomographic imaging with a scintillating glass calorimeter

It has long been suspected that proton therapy has the ability to greatly reduce the dose delivered to healthy tissues in patients compared to traditional forms of radiation therapy, by utilizing the tendency of proton beams to deliver most of their energy (in what is called a Bragg peak) at the end of their path. By strategically placing tumors at the location of a beam’s Bragg peak, the majority of energy is deposited into cancerous tissues. To take advantage of this therapy, accurate proton stopping power images of a patient must be acquired, which is typically done by converting conventional x-ray CT images to stopping power images through a calibration curve. However, this calibration procedure results in non-negligible proton range errors. It has been proposed that this error can be overcome with proton CT imaging, which directly images patient stopping power using protons rather than x-rays. This thesis proposes an inexpensive, compact glass detector that can be attached to existing proton therapy gantries and is capable of performing proton imaging using a specialized image reconstruction algorithm. Additionally, an alternative method of dual-energy CT-based conversion is tested, in order to benchmark the accuracy of the proposed proton imaging scheme by comparing it to a more readily available and competing method-ology. Dual-energy CT combines two different CT images of a patient, utilizing different x-ray sources, to create a composite image of much higher quality than can be achieved with a single scan.