Coronary calcium score scan-based attenuation correction in cardiovascular PET imaging

Abstract

Cardiac positron emission tomography (PET)/CT imaging is a noninvasive procedure allowing the assessment of coronary artery disease (CAD). CT-based attenuation correction of PET data is essential for accurate quantitative analysis in PET/CT imaging. Coronary artery calcium scoring CT (CaScCT) is used as a noninvasive tool for the diagnosis of atherosclerosis in patients with medium risk for CAD. In addition to the CaScCT examination, current cardiac rest/stress NH3 or ¹⁸F-fluorodeoxyglucose viability PET/CT protocols incorporate a correlated low-dose CT scan for attenuation correction purposes (ACCT). As a result, the patient receives a non-negligible radiation dose. The aim of this study is to evaluate the possibility of using CaScCT images for AC of myocardial rest/stress/viability PET data with the aim of reducing patient dose. Since in cardiac PET/CT protocols, the CaScCT examination is usually reconstructed using a small field-of-view, the CaScCT data were reconstructed again with extended field-of-view (ExCaScCT) and used for AC of the corresponding PET data. The feasibility study was performed using 10 patients including four NH3 perfusion and six ¹⁸F-fluorodeoxyglucose viability examinations acquired on the Biograph TP 64 PET/CT scanner. The assessment of PET images corrected using both ACCT and ExCaScCT images was carried out through qualitative assessment performed by an expert nuclear medicine specialist in addition to the regression analysis and the Bland-Altman plots, and 20-segment myocardial bull's eye view analysis. Despite the good agreement between PET images corrected using ACCT and ExCaScCT images as expressed by the correlation coefficient and slope of the regression line in viability (0.949 ± 0.041 and 0.994 ± 0.124) and stress perfusion examinations (0.944 ± 0.008 and 0.968 ± 0.055), the rest perfusion examinations had weak correlation (0.454 ± 0.203 and 0.757 ± 0.193). This is attributed to the fact that the CaScCT scan is performed immediately after the stress/viability ACCT in our protocol that leads to a small misalignment between the CaScCT and stress/viability ACCT images, whereas there is a large misalignment between the CaScCT and rest ACCT images. The bull's eye view analysis showed that the difference between the uptake values was larger in the inferior wall because of diaphragm motion. Our preliminary results seem to suggest that the calcium score study could be used for attenuation correction of cardiac PET images, thus allowing the elimination of ACCT in viability and stress perfusion studies and as such reduce patient dose.