Advanced
Xiaochuan Pan, Ph.D.
Professor
Department of
Radiology and the College
Committee on Medical
Physics
The
Radiation therapy has always been image guided in the
sense that radiation beams are directed at an image-defined target within the
patient. Computed tomography (CT) has been used as the dominant imaging tool for
diagnosis, monitoring, and assessment in image-guided radiation therapy (IGRT).
In the last few years, detector technology has allowed the development of
advanced cone-beam CT for rapid volumetric imaging, and modern cone-beam CT is
playing an increasingly important role in IGRT. The bottleneck of cone-beam CT
imaging had been the lack of adequate theory and algorithms for reconstructing
accurate volumetric images from cone-beam data until the recent theoretic breakthroughs
in cone-beam imaging. In this lecture, following a brief discussion of CT-imaging
basics, I will describe the new developments in cone-beam CT imaging and its
potential applications to IGRT. In particular, I will emphasize the
implications of targeted region-of-interest (ROI) imaging for reduction of
radiation dose, scatter, and other artifacts in repeated imaging in IGRT.
Furthermore, it is always desirable in IGRT practice to reduce CT imaging time
by acquiring, e.g., CT data at a small number of views and/or over a limited
angular range. Accurate image reconstruction from such limited CT data can be
highly challenging. We have recently developed a new approach to addressing
such limited data problems, and I will discuss some interesting results
obtained with this new approach and their potential applications to imaging
tasks in IGRT.