BACKGROUND: Electrocardiography (ECG)-gated scintigraphy demonstrates promising results for the simultaneous assessment of myocardial glucose metabolism and contractile function. In this study a method was evaluated for absolute quantitation of left ventricular wall and cavity parameters with the use of fluorine 18 fluorodeoxyglucose (FDG) ECG-gated positron emission tomography (PET). METHODS AND RESULTS: A previously developed 2-dimensional mathematical model was implemented for computer-automated identification of endocardial and pericardial borders. The accuracy and precision were tested in a heart phantom and in healthy subjects. Twelve healthy men aged 64 +/- 8 years were studied by use of cine magnetic resonance imaging (MRI) and ECG-gated FDG-PET during euglycemic glucose-insulin clamp. At increasing image noise levels, the estimated cavity volume of the heart phantom was within 2 mL of the actual volume, and no significant difference was found between actual and estimated wall thicknesses. Endocardial wall motion as assessed by ECG-gated PET in the healthy subjects was systematically underestimated compared with MRI. This underestimation correlated linearly with endocardial excursion during PET end diastole as measured by MRI. Myocardial end-diastolic wall thickness was systematically overestimated by PET, whereas end-systolic thickness deviated less than 1 mm from MRI. Cavity volumes measured by PET correlated linearly with MRI, with a tendency toward an underestimation of end-diastolic cavity volumes by PET. CONCLUSIONS: Absolute measures of cardiac structure and function may be obtained with a reasonable degree of accuracy by use of ECG-gated PET imaging. However, a high ECG-gating frequency appears to be required to obtain measurements comparable to what may be achieved by MRI.
Journal of Nuclear Cardiology : Official Publication of the American Society of Nuclear Cardiology, 2004, Vol 11, Issue 1, p. 38-46