Dual source dual-energy computed tomography of acute myocardial infarction: correlation with histopathologic findings in a canine model.

PURPOSE: To evaluate the feasibility and value of dual-energy computed tomography myocardial iodine maps in the diagnosis of acute myocardial infarction.

MATERIALS AND METHODS: In 6 dogs, arterial-phase myocardial dual-energy computed tomography imaging were performed 1 day prior to and 3 hours after the surgical ligation of the left anterior descending artery to generate 100 kVp, 140 kVp, average weighted images, and dual energy myocardial iodine maps. For each of the 17 segments of the left ventricle (LV, 102 total segments), the presence or absence myocardial infarction was determined by histopathology and correlated to blinded reader determination of infarcted and noninfarcted myocardium at computed tomography (CT). Statistical analysis for diagnostic accuracy of aforementioned techniques and inter-reader agreement was performed.

RESULTS: The LV myocardial contrast enhancement at the average weighted images and iodine maps were uniform in all 6 dogs before surgery. Following anterior descending artery ligation, histopathology showed 40 infarcted left ventricular segments and 62 noninfarcted segments. For the postligation CT scans, 100 kVp, 140 kVp, average weighted images, and myocardial iodine maps showed 33, 28, 33, 34 infarcted segments and 53, 56, 56, 52 noninfarcted segments for both readers; corresponding to per-segment sensitivities of 83%, 70%, 80%, 92% and specificities of 85%, 90%, 92%, 80% for detecting myocardial infarction. No statistical difference was found for diagnostic accuracy of 100 kV, 140 kV, weighted average images, and iodine maps to detect myocardial infarct segments (all P > 0.05 for both readers). Good inter-reader agreement was seen for myocardial infarct detection using iodine maps (kappa = 0.80).

CONCLUSIONS: Myocardial single- and dual-energy CT imaging shows high per-segment sensitivity and moderate specificity for detecting acute myocardial infarction in a canine model with histopathology as the standard of reference.