Breath-hold CT attenuation correction for quantitative cardiac SPECT
1 Department of Investigative Radiology, National Cerebral and Cardiovascular Center Research Institute, 5-7-1 Fujishirodai, Suita, Osaka 565-8565, Japan
2 Department of Radiology and Nuclear Medicine, National Cerebral and Cardiovascular Center Hospital, 5-7-1 Fujishirodai, Suita, Osaka, 565-8565, Japan
3 Graduate School of Medicine/Faculty of Medicine, Osaka University, 1-1 Yamadaoka, Suita, Osaka, 565-0871, Japan
EJNMMI Research 2012, 2:33 doi:10.1186/2191-219X-2-33Published: 22 June 2012
Attenuation correction of a single photon emission computed tomography (SPECT) image is possible using computed tomography (CT)-based attenuation maps with hybrid SPECT/CT. CT attenuation maps acquired during breath holding can be misaligned with SPECT, generating artifacts in the reconstructed images. The purpose of this study was to investigate the effects of respiratory phase during breath-hold CT acquisition on attenuation correction of cardiac SPECT imaging.
A series of 201Tl-emission and 99mTc-based transmission computed tomography (TCT) scans was carried out along with CT-attenuation scans on 11 young normal volunteers using a hybrid SPECT/CT scanner. The CT scans were performed at three respiratory phases: end-inspiration (INS), end-expiration (EXP), and the midpoint (MID) between these phases. Using alignment parameters between attenuation maps and SPECT images without attenuation or scatter corrections, quantitative SPECT images were reconstructed, including corrections for attenuation and scatter. Regional radioactivity concentrations normalized by the subjects’ weights were compared between CT- and TCT-based attenuation correction techniques.
SPECT images with CT attenuation maps at the EXP phase showed significant differences in regional weight-normalized radioactivity concentrations relative to the images using the other attenuation maps (p < 0.05), as well as systematic positive bias errors, compared to TCT-based images for all myocardial segments, 5.7% ± 2.7% (1.9% to 10.0%). No significant differences in regional weight-normalized radioactivity concentrations were observed between images with CT attenuation maps at MID and INS phases or between these and the TCT-based images, but regional tendencies were found: for anterior to anterolateral segment, positive bias of 5.0% ± 2.2% (1.3% to 8.1%) and 5.6% ± 1.9% (2.6% to 8.5%) and for inferior to inferoseptal segment, negative bias of −5.3% ± 2.6% (−9.1% to −1.7%) and −4.6% ± 2.5% (−8.8% to −1.5%) for the MID and INS phases, respectively.
Use of breath-hold CT attenuation maps at INS and MID phases for attenuation and scatter corrections demonstrated accurate quantitative images that would prove beneficial in cardiac SPECT/CT studies.