Optimizing modelling in iterative image reconstruction for preclinical pinhole PET

Journal article (2016)

Authors

M.C. Goorden RST/Biomedical Imaging - AS
J. van Roosmalen RST/Biomedical Imaging - AS
F. van der Have MILabs B.V., RST/Biomedical Imaging - AS
F.J. Beekman University Medical Center Utrecht, MILabs B.V., RST/Biomedical Imaging - AS
Research Group
RST/Biomedical Imaging (AS) (TU Delft)
Copyright: © 2016 M.C. Goorden, J. van Roosmalen, F. van der Have, F.J. Beekman
DOI
help_outline
https://doi.org/10.1088/0031-9155/61/10/3712
Image reconstruction Positron emission tomography (PET) Single photon emission computed tomography (SPECT)
More Info
expand_more

Published Date

19-04-2016

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Applied Sciences

Department

RST/Radiation, Science and Technology

Research Group

RST/Biomedical Imaging

Abstract

The recently developed versatile emission computed tomography (VECTor) technology enables high-energy SPECT and simultaneous SPECT and PET of small animals at sub-mm resolutions. VECTor uses dedicated clustered pinhole collimators mounted in a scanner with three stationary large-area NaI(Tl) gamma detectors. Here, we develop and validate dedicated image reconstruction methods that compensate for image degradation by incorporating accurate models for the transport of high-energy annihilation gamma photons. Ray tracing software was used to calculate photon transport through the collimator structures and into the gamma detector. Input to this code are several geometric parameters estimated from system calibration with a scanning 99mTc point source. Effects on reconstructed images of (i) modelling variable depth-of-interaction (DOI) in the detector, (ii) incorporating photon paths that go through multiple pinholes ('multiple-pinhole paths' (MPP)), and (iii) including various amounts of point spread function (PSF) tail were evaluated. Imaging 18F in resolution and uniformity phantoms showed that including large parts of PSFs is essential to obtain good contrast-noise characteristics and that DOI modelling is highly effective in removing deformations of small structures, together leading to 0.75 mm resolution PET images of a hot-rod Derenzo phantom. Moreover, MPP modelling reduced the level of background noise. These improvements were also clearly visible in mouse images. Performance of VECTor can thus be significantly improved by accurately modelling annihilation gamma photon transport.