2013 Fiscal Year Final Research Report

Super resolution computed tomography based on Bayesian statistics

Research Project

Project/Area Number	23700574
Research Category	Grant-in-Aid for Young Scientists (B)
Allocation Type	Multi-year Fund
Research Field	Medical systems
Research Institution	The Institute of Physical and Chemical Research
Principal Investigator	JUN Kozuka 独立行政法人理化学研究所, 生命システム研究センター, 研究員 (10432501)
Project Period (FY)	2011 – 2013
Keywords	CT / ベイズ統計 / 超解像
Research Abstract	The reconstruction of a high-quality three-dimensional image from a low-resolution sinogram, which is a visual representation of the two-dimensional projection data obtained from computed axial tomography, is an important problem which arises in fields such as microscope and medical imaging. It is known that several artifacts originated from the inverse Radon transformation arise during typical reconstruction approaches. We have developed a Bayesian treatment of the super-resolution computed tomography problem. This approach is rendered tractable through the introduction of Gaussian processes. Results indicate a significant improvement over techniques based on the filtered back projection.