18 March 2015 Statistical iterative reconstruction for multi-contrast x-ray micro-tomography
Author Affiliations +
Scanning times have always been an important issue in x-ray micro-tomography. To reach high-quality reconstructions the exposure times for each projection can be very long due to small detector pixel sizes and limited flux of x-ray sources. In addition, the required number of projections is a factor which limits a reduction of exposure beyond a certain level. This applies particularly to grating-based phase-contrast computed tomography (PCCT), as several images per projection have to be acquired in order to obtain absorption, phase and dark-field information. In this work we qualitatively compare statistical iterative reconstruction (SIR) and filtered back-projection (FBP) reconstruction from undersampled projection data based on a formalin-fixated mouse sample measured in a grating-based phase-contrast small-animal scanner. The results from our assessment illustrate that SIR offers not only significantly higher image quality, but also enables high-resolution imaging from severely undersampled data in comparison to the FBP algorithm. Therefore, the application of advanced iterative reconstruction methods in micro-tomography entails major advantages over state-of-the-art FBP reconstruction while offering the opportunity to shorten scan durations via a reduction of exposure time per projection and number of angular views.
© (2015) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
S. Allner, A. Velroyen, A. Fehringer, F. Pfeiffer, P. B. Noël, "Statistical iterative reconstruction for multi-contrast x-ray micro-tomography", Proc. SPIE 9412, Medical Imaging 2015: Physics of Medical Imaging, 94123O (18 March 2015); doi: 10.1117/12.2082619; https://doi.org/10.1117/12.2082619

Back to Top