Premium
SU‐D‐116‐04: X‐Ray Transmission Image Reconstruction with Polychromatic Sources for Systems with Nonlinear Response
Author(s) -
Popescu Lucretiu M
Publication year - 2013
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4814057
Subject(s) - iterative reconstruction , image quality , algorithm , iterative method , transmission (telecommunications) , energy (signal processing) , scanner , noise (video) , image resolution , computer science , artificial intelligence , mathematics , image (mathematics) , statistics , telecommunications
Purpose: To develop an image reconstruction algorithm for X‐ray transmission computed tomography (CT) that closely takes into account the physics and statistics of X‐ray emission, transmission and detection, with the aim of producing better quality images. Methods: Based on physical model for X‐ray transmission imaging with polychromatic sources we carry out a data noise analysis applicable to a variety of detector energy‐response functions (e.g. energy integrating, photon counting). From this model we derive multiplicative iterative schemes for the least‐square and the maximum‐likelihood image estimation solutions, and we propose a unified formulation of the two. The same physical and statistical model can also be used for fast generation of X‐ray transmission data at different noise levels. Results: We test the iterative algorithm proposed against the filter back‐projected (FBP) algorithm. The results show that the iterative algorithm using the polychromatic model avoids the cupping artifact due to the beam hardening effect, and the images produced have less noise than FBP images. The signal detectability study, using low contrast signal at unknown locations, shows better performance for the iterative image reconstruction algorithms than FBP. Conclusion: Image reconstruction with detailed modeling of the physics and statistics of the scanner has potential to improve CT image quality.