Date(s) - 03/11/2014
According to the American Society of Radiation Oncology, two–thirds of all cancer patients will receive radiation therapy during their illness with the majority of the treatments been delivered by a linear accelerator (linac). Therefore, quality assurance (QA) procedures must be enforced in order to deliver treatments with a machine in proper conditions. Medical physicists are the professionals in charge of designing, supervising and performing the linac QA tests requiring a tremendous amount of work and hours. A system that can streamline and automate these procedures would be beneficial in aiding the medical physicist in the linac QA part of his/her responsibilities while creating time to improve in other areas such as patient chart checks, consulting and clinical support. Additionally, cost reduction and a program that provides better QA judgment support to the medical physics community are benefits of this project. In summary, automation of a linac QA program will lead to better patient care.
The overall goal of this project is to automate the QA program of a linac. This will be achieved by analyzing and accomplishing various tasks. First, the photon beam dosimetric quantities known as total scatter correction factor (Scp), infinite fractionated depth dose (FDD’) and profiles will be parameterized. Parameterization of the photon beam dosimetry consists of defining the parameters necessary for the specification of a dosimetric quantity model. Currently, these quantities are measured for certain field sizes at specific depths using a three dimensional water tank during annual QA procedures and data collection for linac commissioning. These measurements are susceptible to uncertainties and have various disadvantages including complex equipment setup, extensive procedure involving several professionals introducing interpersonal uncertainties, procedures are not performed regularly decreasing the likelihood of obtaining quality data due to lack of practice, and the amount of required measured data is too large. Parameterizing these quantities has many applications including creating an expert set of data in order to: (1) minimize interpersonal measurement uncertainties, (2) automate the linac QA program, (3) minimize data collection during machine commissioning, (4) implement into the analytical de–convolution method to remove the detector volume averaging effect, and (5) provide a true representation of the beam data for TPS commissioning.
Second, the linac QA tests required by TG–142 will be automated. TG–142 is currently used as a guideline for the linac QA procedures but it has many disadvantages. In summary, its feasibility was not tested, is a long process due to the extensive amount of required tests, it does not address all current technology, and it contains some unnecessary and redundant tasks. The result of analyzing and automating the QA guidelines recommended by TG–142 is a more comprehensible and viable linac QA program that is less time consuming, less error prone and takes into account the newest treatment technologies.
Third, a user–friendly intelligent system for QA automation will be designed. Many professionals entering the field and doing QA procedures are unable to assess if a machine is working properly by observing the collected data and may also use the incorrect measurement setup. Furthermore, time constrains in clinical settings are an added issue. This system will determine machine problems that arise during QA procedures, provide possible solutions to these problems and guide the user through the QA experience. It will assist physicists in performing intelligent judgment in clinics everywhere, save time, prevent machine related accidents and provide better care to the patient.
In conclusion, a comprehensive and streamlined linac QA program will be created from first principle in order to ensure proper functioning of the machine and accurate radiation treatment delivery.