Accuracy of ventricular volume and ejection fraction measured by gated myocardial SPECT: Comparison of 4 software programs

Gated myocardial perfusion SPECT has been used to calculate ejection fracti on (EF) and end-diastolic volume (EDV) and has correlated well with convent ional methods. However, the comparative accuracy of and correlations across various types of gated SPECT software are not well understood. Methods: Ma thematic phantoms of cylindric-hemispheric hybrid models, ranging in volume from 34 to 266 mL, were generated. The clinical cases consisted of 30 pati ents who participated in a radionuclide angiography and gated blood-pool (G BP) study in addition to undergoing Tc-99m-sestamibi gated SPECT. Four kind s of software, Quantitative Gated SPECT (QGS), the Emory Cardiac Toolbox (E CT), 4D-MSPECT, and Perfusion and Functional Analysis for Gated SPECT (pFAS T) were used to compute EF and EDV, and the results were analyzed by multip le comparisons tests. Patients were classified into 4 groups (i.e., no defe ct, small defect, large defect, and small heart) so that factors affecting variation could be analyzed. Results: In mathematic models greater than or equal to 74 mL, volume error was within +/- 15%, whereas for a small volume (34 mL), QGS and 4D-MSPECT underestimated the volume and pFAST overestimat ed it. The respective intra- and interobserver reproducibility of the resul ts was good for QGS (r = 0.99 and 1.00), ECT (r = 0.98 and 0.98), and 4D-MS PECT (r = 0.98 and 0.98) and fair for pFAST (r = 0.88 and 0.85). The correl ation coefficient for EF between gated SPECT and the GBP study was 0.82, 0. 78, 0.69, and 0.84 for QGS, ECT, 4D-MSPECT, and pFAST, respectively. The co rrelation coefficient for EDV between gated SPECT and the GBP study was 0.8 8, 0.89, 0.85, and 0.90, respectively. Although good correlation was observ ed among the 4 software packages, QGS, ECT, and 4D-MSPECT overestimated EF in patients with small hearts, and pFAST overestimated the true volume in p atients with large perfusion defects. Correlation coefficients among the 4 kinds of software were 0.80-0.95 for EF and 0.89-0.98 for EDV. Conclusion: All 4 software programs showed good correlation between EF or EDV and the G BP study. Good correlation was observed also between each pair of quantific ation methods. However, because each method has unique characteristics that depend on its specific algorithm and thus behaves differently in the vario us patient subgroups, the methods should not be used interchangeably.