Fundamental studies of myocardial defect size quantification using positron emission tomography and single photon emission computed tomography

In Flurine-18 fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) acquisition, a transmission scan is usually performed before the PET tracer injection (cold transmission method), followed by a subsequent emission scan. However, this procedure is time consuming. An alternative approach, in which the transmission scan is performed after the emission scan (hot transmission method), would significantly reduce the time required for data acquisition. Recently, three-dimensional PET acquisition (3D PET) has become available. The counting sensitivity is much higher in 3D PET than in conventional two-dimensional PET (2D PET), resulting in a shorter acquisition time and reduced radiation exposure for the patient. On the other hand, 18F-FDG imaging using single photon emission computed tomography (SPECT), a more widely available method than PET, has emerged as an alternative to PET. The purpose of this study was to investigate the accuracy of measurement of myocardial defect sizes by these new techniques, using a chest phantom. Acquisitions were performed using an elliptical cylinder chest phantom. Plastic inserts, ranging in size from 2-60% of the myocardium (n=12), were used as simulated models of transmural myocardial infarction. Fluorine-18 was given into each part of the phantom. PET imaging with cold and hot transmission methods, 3D PET, and SPECT imaging were performed with different acquisition times and different radioisotope concentrations. All PET and SPECT data were analyzed using a semiquantitative polar map approach. Defect sizes were quantified using various cutoff thresholds, and were expressed as a percentage of the left ventricular myocardium. The PET and SPECT measurements were compared with the true defect sizes. Among the various cutoff levels tested, the mean absolute difference between the measured and true defect sizes was minimal at 50% of peak activity for both PET and SPECT. The PET measurements with the hot transmission method showed an excellent correlation with true defect sizes, and the mean absolute error of measurements were similar between the hot and cold transmission methods. In particular, the mean absolute error of 3D PET measurements was smaller than 2D PET for a short acquisition time (10 seconds) and low radioisotope concentrations (25 kBq/ml and 50 kBq/ml). Finally, the SPECT measurements were similar to the PET measurements and closely correlated with the true defect sizes. In conclusion, cardiac PET imaging with these new acquisition techniques is feasible, and 3D PET is more useful than 2D PET in respect of the short acquisition times and low radioisotope concentrations. Furthermore, 18F-FDG SPECT imaging may also be useful to delineate viable myocardium from scarred myocardium in a manner similar to PET. (author)