Volume 1 Supplement 1

Proceedings of the 3rd PSMR Conference on PET/MR and SPECT/MR

Open Access

Iterative reconstruction incorporating positron range correction within STIR framework

  • Ottavia Bertolli1, 2,
  • Matteo Cecchetti2,
  • Niccolò Camarlinghi2,
  • Afroditi Eleftheriou1, 3,
  • Nicola Belcari2 and
  • Charalampos Tsoumpas1
EJNMMI Physics20141(Suppl 1):A42


Published: 29 July 2014

Positron range is one important physical limitation to spatial resolution in PET imaging. We present a method to take account of positron range blurring in iterative reconstruction by including specific positron range kernels approximating the annihilation distribution in the forward projection operation [1]. The correction method is implemented within the OSMAPOSL iterative reconstruction algorithm of STIR 3.0 [2].

GATE [3] simulation package was used to obtain the positron annihilation distribution of a point source of 68Ga placed in the centre of spherical water and lung tissue phantoms. The simulations were performed without and with 3T static magnetic field in the axial direction. To obtain the kernels the annihilation coordinates were voxelized into a 3D matrix with element size equal to the reconstructed image's voxel size and then normalised. The proposed method was evaluated using simulation data (generated with STIR) of a parallelepiped water phantom containing two spherical hot spots, on the HYPERImage preclinical PET/MR scanner [4]. To simulate the positron range effect the phantom was blurred by convolution with the positron kernel relative to water. The blurred images were reconstructed with and without filtering of the forward projected image.

Line profiles traced through the spots of the reconstructed images show that the use of the positron range correction yields sharper boundary definition, albeit with noise enhancement. The values relative to the spots volume show that the correction results in the recovery of the activity mean value compared to the simulated phantom (e.g. 55% to 94% for the smaller spot, no magnetic field).

Furthermore the tissue-dependent correction, which consists in blurring the image with the kernel correspondent to the voxel tissue (water or lung), was implemented and evaluated on a phantom with water and lung regions. The activity mean value is recovered but the method performance in correspondence to the edges needs further investigation.

Authors’ Affiliations

Division of Medical Physics, University of Leeds
Department of Physics, University of Pisa and INFN Pisa
Department of Physics, National and Kapodistrian University of Athens


  1. Cal-Gonzalez J, Herraiz JL, et al.: Positron range effects in high resolution 3D PET imaging. IEEE Nucl. Sci. Symp. Med. Imag. Conf. Rec 2009, 2788–91.Google Scholar
  2. Thielemans K, et al.: STIR: software for tomographic image reconstruction release 2. Phys Med Biol 2012,57(4):867–883. 10.1088/0031-9155/57/4/867PubMedView ArticleGoogle Scholar
  3. Jan S, et al.: GATE: a simulation toolkit for PET and SPECT. Phys Med Biol 2004, 49: 4543–4561. 10.1088/0031-9155/49/19/007PubMed CentralPubMedView ArticleGoogle Scholar
  4. Schulz V, et al.: A preclinical PET/MR insert for a human 3T MR scanner. IEEE Nucl. Sci. Symp. Med. Imag. Conf. Rec. 2009, 1–5: 2577–2579.Google Scholar


© Bertolli et al; licensee Springer 2014

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.