In SPECT myocardial perfusion imaging (SPECT MPI) attenuation correction by means of CT (CT-AC) is mandatory to reduce attenuation artifacts. On the other hand CT-AC could itself introduce substantial artifacts. Aim of this study was to use a, properly filled, anthropomorphic torso phantom to check the quality of CT-AC. Qualitative and semi-quantitative analysis of truncation artifacts was performed after serial SPECT/CT acquisitions at different FOVs (field of views).
We adopted the “Anthropomorphic Torso Phantom” (Mod. ECT/TOR/P – Data Spectrum Corporation) divided in four regions: two lungs (filled with Styrofoam and water to achieve a density similar to lung), a thoracic area and a liver area. A static cardiac insert simulating left ventricule was also inserted in the torso phantom. Over the Torso phantom two different size of breasts (Large Solid Breast Accessory Set, Model ECT/SOL-BR/L) phantom were used to vary attenuation. Gamma camera used was General Electric Medical System mod. Infinia II
The Myovation Protocol (Xeleris – GE) was used to compare the attenuation corrected images and bull’s eye image with the uncorrected ones. Truncation artifacts originated by increasing zoom (zoom 1.0, 1.6 or 2.5) were analyzed respectively in the female and in the male phantom.
Surprisingly, we observed that (for the combination of tested gamma camera and software) only the CT portion of the entire CT-AC corresponding to the emissive data was used for the correction. The remainder CT portion (that increases with zoom) was discarded by software resulting in underestimation of AC and corresponding artifacts. In consequence, a zoom acquisition factor equal to 1.0, which incudes the majority of the anatomy, is mandatory in the tested setting especially when dealing with large breast and obese patients. Zoom larger than 1 (1.6 and 2.5) revealed increasing truncation artifacts in lateral side of the phantom.
Before reliably trusting a vendor provided CT-AC, it is very useful to test results with the aid of advanced quality controls and phantoms to be self-assured that no artifacts are introduced by the attenuation correction. This is also important to fine tune acquisition parameters to ensure maximum resolution and sensibility but avoiding truncation.