UNLABELLED: With the introduction of 99mTc-teboroxime, a previously undocumented artifact has shown up in cardiac SPECT imaging. In the images, the uptake values near the inferior wall are lower than expected. The artifact has been reported in the literature, but an adequate explanation has not yet been provided. The high uptake of 99mTc-teboroxime in the liver has been demonstrated to be the cause of this artifact. METHODS: With simulations we show that an artifact can be reproduced by applying filtered backprojection (without corrections for attenuation) of attenuated and blurred projections. The conclusions from the simulations are validated with SPECT and PET phantom measurements. Maximum likelihood expectation maximization (ML-EM) reconstruction is applied to evaluate the effect of accurate attenuation correction. The influence of the high liver uptake on the convergence of ML-EM was also evaluated. RESULTS: The artifact results mainly when the photon attenuation during reconstruction is ignored. This results in a distorted reconstruction of the liver. These distortions affect the neighboring inferior wall of the myocardium. While the use of opposite projections reduces the effect, accurate attenuation correction nearly eliminates it. A small additional deformation is caused by the position dependence of the spatial resolution of the gamma camera. It was also noted that the presence of the liver slows down the convergence of ML-EM in the heart region. CONCLUSION: The liver-heart artifact is an attenuation effect and is eliminated by attenuation correction. The local convergence of ML-EM is affected by the total image content.
UNLABELLED: With the introduction of 99mTc-teboroxime, a previously undocumented artifact has shown up in cardiac SPECT imaging. In the images, the uptake values near the inferior wall are lower than expected. The artifact has been reported in the literature, but an adequate explanation has not yet been provided. The high uptake of 99mTc-teboroxime in the liver has been demonstrated to be the cause of this artifact. METHODS: With simulations we show that an artifact can be reproduced by applying filtered backprojection (without corrections for attenuation) of attenuated and blurred projections. The conclusions from the simulations are validated with SPECT and PET phantom measurements. Maximum likelihood expectation maximization (ML-EM) reconstruction is applied to evaluate the effect of accurate attenuation correction. The influence of the high liver uptake on the convergence of ML-EM was also evaluated. RESULTS: The artifact results mainly when the photon attenuation during reconstruction is ignored. This results in a distorted reconstruction of the liver. These distortions affect the neighboring inferior wall of the myocardium. While the use of opposite projections reduces the effect, accurate attenuation correction nearly eliminates it. A small additional deformation is caused by the position dependence of the spatial resolution of the gamma camera. It was also noted that the presence of the liver slows down the convergence of ML-EM in the heart region. CONCLUSION: The liver-heart artifact is an attenuation effect and is eliminated by attenuation correction. The local convergence of ML-EM is affected by the total image content.
Authors: Irene A Burger; David A Scheiner; David W Crook; Valerie Treyer; Thomas F Hany; Gustav K von Schulthess Journal: Eur J Nucl Med Mol Imaging Date: 2010-09-21 Impact factor: 9.236
Authors: Kengo Hatada; Mirta Ruiz; Laurent M Riou; Ronaldo L Lima; Allen R Goode; Denny D Watson; George A Beller; David K Glover Journal: J Nucl Cardiol Date: 2006-11 Impact factor: 5.952
Authors: M A King; W Xia; D J deVries; T S Pan; B J Villegas; S Dahlberg; B M Tsui; M H Ljungberg; H T Morgan Journal: J Nucl Cardiol Date: 1996 Jan-Feb Impact factor: 5.952