Myocardial perfusion imaging and coronary calcium scoring with a two-slice SPECT/CT system: can the attenuation map be calculated from the calcium scoring CT scan?

Abstract

Coronary artery calcium scoring can complement myocardial perfusion imaging (MPI). The purpose of this study was to evaluate the feasibility and accuracy of using the CalciumScore-CT derived from a combined SPECT/CT device also for SPECT attenuation correction (AC). The study group comprised 99 patients who underwent both post-stress and rest MPI using a two-slice SPECT/CT system. For AC, one of the two scans was accompanied by a CalciumScore-CT scan (CalciumScore-CTAC) and the other by a conventional spiral CT (AttenCorr-CT) scan (AttenCorr-CTAC). In 48 patients the CalciumScore-CT scan was acquired with the post-stress scan and the AttenCorr-CT scan with the rest scan, and in 51 patients the order was reversed. The accuracy of the images based on AC was determined qualitatively by consensus reading with respect to the clinical diagnoses as well as quantitatively by comparing the perfusion summed stress scores (SSS) and the summed rest scores (SRS) between attenuation-corrected and uncorrected images. In comparison to the uncorrected images CalciumScore-CTAC led to regional inaccuracies in 14 of 51 of studies (27.5%) versus 12 of 48 studies (25%) with AttenCorr-CTAC for the stress studies and in 5 of 48 (10%) versus 1 of 51 (2%) for the rest studies, respectively. This led to intermediate and definite changes in the final diagnosis (ischaemia and/or scarring) in 12% of the studies (12 of 99) and in 7% of the studies (7 of 99) with CalciumScore-CTAC and in 9% of the studies (9 of 99) and 4% of the studies (4 of 99) with AttenCorr-CTAC. Differences in SSS and SRS with respect to the uncorrected images were greater for the CalciumScore-CTAC images than for the AttenCorr-CTAC images (ΔSSS 4.5 ± 5.6 and 2.1 ± 4.4, p = 0.023; ΔSRS 4.2 ± 4.9 and 1.6 ± 3.2, p = 0.004, respectively). Using the same CT scan for calcium scoring and SPECT AC is feasible. Image interpretation must, however, include uncorrected images since CT-based AC relatively often introduces artefacts into the myocardial perfusion images. This effect is somewhat more pronounced with CalciumScore-CTAC than with AttenCorr-CTAC.