The basic goal of radiotherapy is to expose cancer cells of a patient to lethal doses of radiation, while sparing the surrounding healthy tissues. Even though this goal appears simple, challenges are many. Modern radiation therapy has evolved into a complex process. Linear accelerators and treatment planning systems are pushed to their limits trying to deliver highly conformal dose of radiation. The risks and potential effects of errors therefore increase with the newest forms of treatment. Improving precision and controlling the quality of radiation treatments at every step of the process, even during delivery, is vital. Patient survival ultimately depends on the quality and precision of the treatment
The proposed work deals with online, in vivo, verification of radiation treatments. This verification will be carried in parallel on two fronts: dose verification and anatomy tracking. First, we will develop miniature, unobtrusive, radiation detectors to perform a real-time, in vivo, monitoring of radiation doses. Second, we will automate anatomy tracking by developing algorithms to integrate information from multiple image sources (planning CT, cone beam CT, radiographs) and combining them through deformable registration. By uniting the dose information with the anatomical information acquired, we will be able to identify patients who risk a deviation from their planned treatment objectives. We will then deploy a gradual set of intervention strategies to correct this situation. Using a gradual approach starting with relatively unobtrusive steps will allow us to monitor large number of patients with minimal impact on the clinical workflow.
Implementation of this treatment quality verification process will lead to better, more personalized, radiation treatment for patients who are most susceptible to benefits from them. In the long term, better precision and accuracy will facilitate dose escalation to the cancer without increased risks for the healthy tissues.