Tomographic Imaging from Limited Cone-Beam Data In several diverse areas of cancer diagnosis and treatment, new imaging applications could be made available if volumetric tomographic imaging were possible from highly irregular and unconventional projection data configurations. One major difficulty is finding reliable reconstruction algorithms. This proposals describes an approach to address the following scientific questions: What features can and can not be tomographically reconstructed from a specified limited data configuration? Can algorithms be devised to handle various limited data situations without producing subtle or unpredictable limited-data artifacts? The specific aims of this research project are: 1. To develop a theory linking limited cone-beam (CB) projection data to achievable reconstructed image quality; and to design and implement an algorithm. to quantitatively describe the maximum tomographic capability from a given limited data configuration. 2. To devise, implement, and test new algorithms for tomographic reconstruction from limited CB data configurations for practical imaging, and to demonstrated tomographic reconstructions from stimulated data in two specific limited-data imaging applications.
The aims will be achieved by performing a sequence of experiments to refine the theoretical models and algorithm developments. The experimental work will use computer simulations. The principles will be demonstrated with simulations of two cancer-related applications: tomographic imaging of tracers for diagnostic breast imaging using pinhole collimators; and tomographic imaging in the interventional operating suite using fluoroscopy projection images.
Noo, Frederic; Defrise, Michel; Clackdoyle, Rolf et al. (2002) Image reconstruction from fan-beam projections on less than a short scan. Phys Med Biol 47:2525-46 |