Date of Award
Doctor of Philosophy
Many portal imaging devices have been developed to verify the geometric accuracy of radiation therapy treatments. Portal imaging devices are used to take images of the patient during radiation therapy treatments. These images are used to detect patient positioning errors which may jeopardize the outcome of conventional and high-precision radiotherapy treatments. Unfortunately, the quality of portal images obtained with such devices is disappointing, resulting in sparse clinical use of these devices.;Researchers have been substituting various imaging components on these portal imaging systems in the hopes of optimizing portal image quality. This empirical approach has led to some successes. However, choosing imaging system components on the basis of one desirable parameter while ignoring the impact of the change on overall system performance wastes time, money, and effort. Clearly, a more efficient approach is required.;This thesis presents approaches for the optimization of both the design and use of portal imaging devices. These approaches require understanding of the fundamental physics of portal imaging, such as the size and shape of the x-ray sources of medical linear accelerators and the interaction of x-rays within typical portal imaging detectors. The use of existing portal imaging systems (i.e., portal films and video-based systems) has been optimized by finding the radiographic magnification which provides the best image quality for a particular system/linear accelerator combination. It has been found that, for portal films, radiographic magnification is undesirable. On the other hand, a radiographic magnification of 1.5-1.7 is optimal for video-based systems. Therefore, the image quality from an existing imaging system can be improved without changing the system design. The design of portal imaging systems has been optimized using a theoretical approach known as quantum accounting diagram (QAD) theory. Using this theory, a detailed analysis of a video-based portal imaging system has permitted the theoretical derivation of the detective quantum efficiency (DQE) of the imaging system. The analysis has shown that the video-based portal imaging system suffers from severe secondary quantum sinks for non-zero spatial frequencies, resulting from sub-optimal system design. Furthermore, the theoretical DQE's have been compared with experimental measurements--the first experimental verification of the QAD theory. The QAD theory has been expanded to include the physical parameters involved with the human visual system and allow the computation of indices of perceived image quality. This approach enables the optimization of imaging devices using a single figure of merit, and has been used to optimize the phosphor screen thickness used by two different designs of portal imaging systems. We have demonstrated that the QAD approach allows us to predict the change in overall system performance for any modification in imaging system design. In future, we believe that this tool will be vital to the development and optimization of improved portal imaging systems for radiation therapy.
Bissonnette, Jean-pierre, "Optimization Of Portal Imaging Systems" (1996). Digitized Theses. 2681.