AJR:195, September 2010
Ralf W.Bauer, J. Matthias Kerl, Nadine Fischer, Thorsten Burkhard, Maya C. Larson,1 Hanns Ackermann, Thomas J. Vogl
OBJECTIVE. The purpose of this article is to compare the performance of dual-energy CT with that of 3-T MRI with late enhancement for the detection of chronic myocardial in¬farction during first-pass coronary CT angiography (CTA).
SUBJECTS AND METHODS. Thirty-six patients underwent coronary CTA for the assessment of coronary bypass graft patency on a first-generation dual-source CT scanner in dual-energy mode. Gray-scale images (100 kV, 140 kV, and blended virtual 120 kV) were assessed for areas of hypodense myocardium during the arterial phase. In addition, a color-coded map of myocardial iodine distribution was calculated from the dual-energy data for perfusion analysis. Dual-energy CT data were compared with data from 3-T MRI with late enhancement, which served as the reference standard for scar detection using the American Heart Association's 17-segment model of the left ventricle.
RESULTS. One hundred one (17%) of 612 myocardial segments in 22 (61%) of 36 patients showed late enhancement on MRI. Although myocardial iodine mapping was prone to artifacts, mostly arising from sternal wires (70% sensitivity), 100-kV gray-scale images showed the high¬est sensitivity (80%) for the detection of myocardial scar. Blended virtual 120-kV images with lower noise and higher resolution had the best diagnostic accuracy (77% sensitivity, 97% speci¬ficity, 85% positive predictive value, 96% negative predictive value, and 94% accuracy).
CONCLUSION. Detection of chronic myocardial infarction on color-coded iodine dis¬tribution analysis with first-generation dual-energy CT is impeded by thoracic metallic devic¬es. This group of patients benefits more from adequate blending of high- and low-kilovoltage gray-scale images. Further technical improvements are desirable to lower artifact burden and improve sensitivity on myocardial iodine distribution mapping.