Evaluation of Myocardial SPECT Imaging Reconstructed from 270°Projection Data

—A Study Using a Cardiac Phantom—

Abstract

SPECT reconstruction is commonly performed using 360°or 180°projection data. However, it is also possible to reconstruct SPECT images using other projection data arcs. The purpose of this study was to characterize images obtained by limiting the projection data to 270°by discarding the projection views with severe attenuation (Fig. 1) . A series of phantom studies was performed with and without plastic chambers simulating perfusion defects using ²⁰¹Tl and ^99mTc. Images using 270°, 360°, and 180°projection arcs were identically reconstructed from the same data. In the absence of plastic chambers, intraslice uniformity in a given slice was assessed by computing the coefficient of variation (CV) of average counts in 8 ROIs within the slice. Interslice uniformity was assessed by computing the CV of average counts in five short axial slices. With plastic chambers in place, the variability in defect contrasts was assessed by computing the CV of defect contrasts in 4 chambers, located on the anterior, lateral, inferoposterior, and septal walls. The intraslice uniformity of the 270°images were considerably inferior to those of the 360°and 180° images. The interslice uniformity was highest in the 360°images, and lowest in the 180°images. The variation in defect contrasts in the 270°image was higher than those of the other two images. The 270°images showed a high defect contrast in the septum and high counts in the anterior and anteroseptal wall. Because a large variation in defect contrasts within a segment might result in false positive or negative in diagnosis, 270°imaging is not recommended over 360°or 180°imaging.