ABSTRACT
Clinical outcome studies with clear and objective endpoints are necessary to make informed radiotherapy treatment decisions. Commonly, clinical outcomes are established after lengthy and costly clinical trials are performed and the data are analyzed and published. One the challenges with obtaining meaningful data from clinical trials is that by the time the information gets to the medical profession the results may be less clinically relevant than when the trial began, An alternative approach is to estimate clinical outcomes through patient population modeling. We are developing a mathematical tool that uses Monte Carlo techniques to simulate variations in planned and delivered dose distributions of prostate patients receiving radiotherapy. Ultimately, our simulation will calculate a distribution of Tumor Control Probabilities (TCPs) for a population of patients treated under a given protocol. Such distributions can serve as a metric for comparing different treatment modalities, planning and setup approaches, and machine parameter settings or tolerances with respect to outcomes on broad patient populations. It may also help researchers understand differences one might expect to find before actually doing the clinical trial. As a first step and for the focus of this abstract we wanted to see if we could answer the question: "Can a population of dose distributions of prostate patients be accurately modeled by a set of randomly generated Gaussian functions?" Our results have demonstrated that using a set of randomly generated Gaussian functions can simulate a distribution of prostate patients.
ABSTRACT
Improved treatment techniques in radiation therapy provide incentive to reduce treatment margins, thereby increasing the necessity for more accurate geometrical setup of the linear accelerator and accompanying components. In the present paper, we describe the development of a novel device that enables precise and automated measurement of geometric parameters for the purpose of improving initial setup accuracy, and for standardizing repeated quality control activities. The device consists of a silicon photodiode array, an evaluation board, a data acquisition card, and a laptop. Measurements that demonstrate the utility of the device are also presented. Using the device, we show that the radiation light field congruence for both 6 and 15 MV beams is within 1.3 mm. The maximum measured disagreement between radiation field edges and light field edges was 1.290 ± 0.002 mm, while the smallest disagreement between the light field and radiation field edge was 0.016 ± 0.003 mm. Because measurements are automated, ambiguities resulting from interobserver variability are removed, greatly improving the reproducibility of measurements across observers. We expect the device to find use in consistency measurements on linear accelerators used for stereotactic radiosurgery, during the commissioning of new linear accelerators, or as an alternative to film or other commercially available devices for performing monthly or annual quality control checks.
