Development of Prior Image-based, High-Quality, Low-Dose Kilovoltage Cone Beam CT for Use in Adaptive Radiotherapy of Prostate Cancer.

摘要

Adaptive radiotherapy (RT) is an advanced technique for prostate cancer treatment which employs kilovoltage (KV) conebeam CT (CBCT) for guiding treatment. High quality CBCT images are important in achieving improved treatment effect, but they also require a non-negligible amount of imaging radiation dose which raises patient safety concern. Therefore, the goal of this project is to investigate and develop innovative, prior-image-based CBCT imaging techniques that can yield high quality images with reduced dose. During the second year of this project, I have investigated and implemented novel scanning trajectories with aid of additional hardware components and software control. I carried out a rigorous investigation on developing optimization-based image reconstruction algorithms by specifying a discrete imaging model, formulating optimization programs, and designing and implementing iterative algorithms. Prior images have been investigated for incorporation in image reconstruction tasks for all the data acquisition configurations. Quantitative metrics have been calculated for quantitatively characterizing and evaluating the reconstruction quality, choice of parameters, and strategies for incorporating prior images. The research that I carried out during the second year was as planned in the proposal, and the results built a solid basis for continuing further investigation planned for the next year.