Reproducibility and quantitativity of oblique-angle reconstruction in single photon emission computed tomography using Tl-201 myocardial phantom

This study was carried out in order to evaluate reproducibility and quantitativity of oblique-angle reconstruction of myocardial phantom SPECT. Myocardial phantom with transmural and subendcardial defects, and off-axis phantom with wall thickness changing continuously from 0 to 23 mm were used. Sixty projection data in every 60 were aquired using dual-camera (ZLC) with high resolution collimators connected to Scintipac-2400 computer system. Oblique-angle reconstructed images were obtained by indicating the long axis of the phantom manually in the transaxial and vertical long axial tomograms. Reproducibility and quantitativity were evaluated by creating circumferential profile (CFP) of the finally reconstructed short axial images. Inter- and intra-operater reproducibility of relative counting ratio were less than 6.7% (C.V.) and 3.3% (C.V.), respectively. Both inter- and intraoperater reproducibility of absolute counts were better than that of counting ratio (less than 5.1% (C.V.) and 2.9% (C.V.), respectively). Variation of defect location in the reconstructed image and between the slices were less than 1 sampling interval of CFP (60) and 0.6 slice, respectively. Quantitativity of counts in the reconstructed images was poor in the transmulal defect, but was fair in the subendocardial defect. Counting ratio was greatly affected by wall thickness. Temporal quantitatibity or linearity of the counts in sequential SPECTs was good in non-defect area, especially when wall thickness was greater than 70% (16 mm) of maximum. In conclusion, three-dimensional oblique-angle reconstruction in Tl-201 myocardial SPECT could be applicable to relative and temporal quantitation of local myocardial activity other than defect area for the quantitative evaluation of Tl-201 myocardial wash-out. (J.P.N.)