Artifacts from implanted leads in cardiac PET using CT-based attenuation correction

Abstract

Conclusion: The LV EF and volumes are comparable in patients with and without type II DM in the absence of ischemia or infarction. The LV volumes and EF are lower in women than men in the presence or absence of DM. It is possible that ischemia/infarction affect more adversely the LV function in DM accounting for the worse LV function [and worse outcome] in patients with established CAD. Background: The use of CT-based attenuation correction (CT-AC) in cardiac PET can potentially cause artifacts, not just due to respiratory phase mismatch, but also due to implanted metallic leads. The significance of such metallic lead artifacts requires investigation. Methods: PET images of phantoms containing pacemaker leads and ICD leads in uniform FDG solution were acquired and analyzed. The leads were mounted in 90-degree bent configurations, allowing the magnitude of the artifact to be assessed when the lead is normal to or parallel to the transaxial planes. FDG cardiac PET studies of patients having implanted pacemaker (n=4) or ICD leads (n=4) were also analyzed. In all cases, images were acquired on both dedicated PET and PET-CT scanners, to compare differences of transmission sources vs. reduced-dose 120kVp “slow” CT for attenuation correction. Results: The magnitude of artifact depends on several factors: composition, design, and location. Unlike internal conductor wires (of low atomic number metal), external platinum alloy electrodes are much more likely to cause artifacts. Pacemaker leads, which have relatively small electrodes, caused noticeable artifacts in CT images but not in the resultant CT-AC PET or the dedicated PET images. ICD leads, which have much larger electrodes (shock coils), did cause significant regional artifacts in CT-AC PET images, but not in dedicated PET images. The apparent increase in FDG concentration from ICD metal artifact was on the order of 80% for phantom images (with the lead oriented in-plane) and 30% in patient myocardial images. Clinically significant artifacts were observed when the ICD lead was implanted adjacent to the left ventricle, but not when adjacent to the right ventricle. Conclusions: ICD leads can produce clinically significant artifacts in CT-AC cardiac PET, depending on their position and orientation, whereas pacemaker leads do not. Automated methods to correct for metal artifact would be beneficial, since the prevalence of implanted ICDs in patients undergoing cardiac PET is likely to rise. These results also may have implications for CT-AC cardiac SPECT imaging.