Radiographic Proton Imaging Using the Mevion-S250 Radiation Therapy System with Gafchromic Films and Computed Radiograph

Image-guided radiation therapy (IGRT) with on-board imaging systems provides updated images of patient anatomy using 2D-radiographic and 3D-CT imaging tools. Images from on-board imaging systems for IGRT are used to setup patients based on internal patient anatomy in most of the institutions and to evaluate patient anatomical variation just before dose delivery. IGRT allows accurate patient setup with smaller margins and lower doses for normal tissue and critical structures. Volumetric cone-beam CT images from IGRT with updated patient anatomy from images used for initial treatment planning can be used to consider anatomical variations and re-plan certain cases with adaptive radiation therapy. In this project, the image quality and dose of radiographic images with therapeutic 250 MeV proton beams from the MEVION-S250 system using Gafchromic films and computed radiographic plates were investigated quantitatively. Proton beams from the MEVION-S250 machine were used to acquire radiographic images of different phantoms that include the Leeds, Catphantom, Las Vegas, anthropomorphic head, and pelvis phantoms. The MEVION system provided a double scattering 250-MeV proton beam shaped with two nozzles (14 and 25cm diameters) and various beam ranges (5--32 cm) and range modulation (2--20 cm). The different testing objects in the Leeds phantom were used to quantify image quality. Imaging quality dependence on dose was quantified using different dose levels (0.1--12 MU). The Leeds phantom was place upstream outside the Bragg peak range while the image receptor either of films or CR-cassettes were placed downstream to measure the exiting protons. Image quality of the proton radiographic images was affected by multiple factors that include: (a) depth in phantom, (b) dose or monitor units (MU) from the proton imaging beam, (c) separation between the phantom and image receptor, (d) beam modulation and (e) beam range. Despite difficulties inherent to proton imaging, it has potential advantages particularly in proton treatment. If a proton beam is used for imaging in a way such that the Bragg Peak were deeper than the object being imaged, the imaging doses could potentially be lower compared to the imaging doses of traditional photon imaging techniques. Proton imaging may also improve proton dose calculations for proton treatment planning in addition to patient positioning accuracy. Image quality of the Leeds phantom improved with depth in phantom as the phantom location approached the Bragg-peak where the best position and contrast resolutions were obtained. As dose increased, image quality improved with better signal to noise ratio particularly at the Bragg-peak. The increased separation between the Leeds phantom and imaging plates leads to increased scatter or bending of the proton beam which degraded the image quality of the proton radiographic images. Short beam ranges and wider modulations improved image quality as the phantom became closer to or within the spread-out Bragg-peak, respectively. In conclusion, this study investigated quantitatively the dependence of image quality on different parameters including phantom depth, separation between the phantom and image receptor, range and modulation of high energy therapeutic proton beam. Radiographic proton images were feasible for clinical applications and had comparable image quality to kV-diagnostic photon images. The imaging doses deposited from the proton beams were comparable to or up to 15% lower than the imaging dose deposited by clinically used photon imaging techniques used for image-guided radiation therapy from the on-board imaging system.