End-to-End Quality Assurance of Complex Radiation Therapy Treatments

Abstract

Radiation treatments have become increasingly complex, making quality assurance (QA) of radiation therapy deliveries of vital importance. In order to measure complex dose deliveries, it is crucial to have the proper dosimetric, readout, and analysis tools that can accurately report the delivered dose. While clinics generally perform a wide array of quality assurance testing on specific steps of the planning and delivery of the radiation, they seldom perform routine end-to-end testing of the entire treatment planning and delivery process. Such end-to-end tests are a valuable QA tool to help ensure that all individual components are working together to deliver the planned dose. Using an anthropomorphic phantom that incorporates several dosimeters (with fiducials for localization) for end-to-end QA can help to mimic a patient treatment and to study the overall geometric and dosimetric accuracy of an entire treatment process.

In this thesis, we look to perform comprehensive end-to-end quality assurance of radiation therapy treatments using point detectors, and two-dimensional (2D) and 3D dosimeters. Specifically, this work examines: (i) improving film dosimetry readout methods by testing the capabilities of our inhouse diffuse light field and camera film imaging system, as well as developing a new point source film scanner to eliminate stray light issues present in the diffuse light system, (ii) developing a streamlined gel dosimetry workflow and validating new gel dosimetry analysis software, (iii) using multiple dosimeters, coupled with commercial and in-house developed phantoms, to inform and verify clinical protocols and to commission new treatment techniques, (iv) characterizing and benchmarking various 3D printing materials (for applications in 3D printed patient bolus and custom phantom manufacturing), (v) and developing an anthropomorphic head phantom that accommodates multiple dosimeters, using it to test a complete clinical treatment process from beginning to end (exactly like an actual patient would be treated), and involving clinical staff at each step of the process.