Image improvement in pinhole SPECT using complete data acquisition combined with statistical image reconstruction

Tsutomu Zeniya, Hiroshi Watabe, Toshiyuki Aoi, Kyeong Min Kim, Noboru Teramoto, Takuya Hayashi, Antti Sohlberg, Hiroyuki Kudo, Hidehiro Iida

Research output: Contribution to journalArticlepeer-review


Pinhole single-photon emission computed tomography (SPECT) with high spatial-resolution is suitable for small-animal imaging, but has limitations associated with spatial-resolution inhomogeneity or axial blurring. We have hypothesized that this blurring is due to incompleteness of projection data acquired by a single circular pinhole orbit. And we have developed a pinhole SPECT system with two circular orbits which satisfy Tuy's condition so as to provide complete data for 3D pinhole SPECT reconstruction within the whole field-of-view (FOV); a dedicated 3D ordered subsets expectation maximization (3D-OSEM) reconstruction method for two-orbit data. This study is aimed at evaluating accuracy and impact of this system. In this system, not the camera but the object rotates and the two orbits are 90° and 45° relative to object's axis. Experiments using a multiple-disk phantom filled with 99mTc solution and a mouse bone scan using 99mTc-labeled HMDP agent were carried out. The Feldkamp's filtered back-projection (FBP) method and the 3D-OSEM method were applied to these data sets. The axial blurring was apparent on images reconstructed by FBP for single-orbit data, while the 3D-OSEM using two-orbit data dramatically improved the resolution homogeneity and statistical noise property, and also demonstrated considerably better image quality in the mouse scan.

Original languageEnglish
Pages (from-to)101-105
Number of pages5
JournalInternational Congress Series
Issue numberC
Publication statusPublished - 2004 Aug 1


  • Complete data acquisition
  • Pinhole SPECT
  • Small animal
  • Statistical image reconstruction
  • Tuy's condition


Dive into the research topics of 'Image improvement in pinhole SPECT using complete data acquisition combined with statistical image reconstruction'. Together they form a unique fingerprint.

Cite this