J Nucl Med. 2007; 48 (Supplement 2):420P
Instrumentation & Data Analysis: Image Generation
Image Generation Posters |
Comparison of attenuation-correction artefacts in cardiac imaging using PET/CT and SPECT/CT
Sarah Martin1 and
Brian Hutton1
1 Institute of Nuclear Medicine, University College London, London, United Kingdom
1753
Objectives: Respiratory motion has been shown to cause artefacts in myocardial perfusion imaging in PET/CT due to mismatches arising between the emission and transmission acquisitions. However the artefacts occurring in SPECT/CT from this same source have not been demonstrated. Our goal was to make a direct comparison of the errors due to respiratory-induced attenuation mismatches in PET/CT and SPECT/CT. Methods: Activity maps at specific respiratory phases were produced using the NCAT software phantom, and were projected and reconstructed to produce SPECT and PET images using appropriate projectors for each modality. Attenuation maps at mismatched respiratory phases were used for the attenuation-correction and the effects of mismatches on the subsequent images were assessed. This was carried out for different types of mismatch (cardiac translation, lung expansion due to diaphragm contraction and lung expansion due to chest wall motion), all of which are involved in normal breathing. This enabled the causes of artefact to be determined and compared for SPECT versus PET. The severity of artefacts was assessed by recording the average values in 10 regions-of-interest in the mid-slices of the vertical and horizontal long axes and calculating the variation in these averages. Results: The measured variations in the images produced are listed in the following table. [table] Conclusions: In general, PET and SPECT images were affected in a similar manner by attenuation mismatches, but with a more pronounced effect in PET. In both modalities attenuation mismatches due to cardiac translation lead to the largest changes, with PET being more severely affected. Lung expansion due to chest wall motion did not increase the variation in either modality, whereas lung expansion due to diaphragm contraction caused an increase only in PET images. The need for attention to attenuation mismatch is well recognised in PET/CT and these results suggest similar attention needs to be paid to mismatch in SPECT/CT with similar correction methods potentially applicable.
