Comparison of software programs for the assessment of left ventricular ejection fraction using 99mTc-tetrofosmin-gated SPECT/CT: correlation with equilibrium radionuclide ventriculography in the Indian population.

OBJECTIVE: The goal of this study was to compare Emory Cardiac Toolbox (ECTb), quantitative gated SPECT (QGS), four-dimensional single photon emission computed tomography (4D-MSPECT) and Myometrix cardiac software programs for the assessment of left ventricular ejection fraction (LVEF) using 99mTc-tetrofosmin-gated SPECT/CT [myocardial perfusion SPECT (MPS)] and correlate them with the LVEF values derived from equilibrium radionuclide ventriculography (ERNV) in patients with known/suspected coronary artery disease (CAD).

MATERIALS AND METHODS: A total of 109 patients (80 men, 29 women) were recruited into the study. Fifty-five patients had known CAD and 54 were referred with suspicion of CAD. All the patients underwent ERNV and MPS as per the standard protocol. ERNV was processed using the vendor-provided 'EF analysis' and gated MPS was processed using individual software programs.

RESULTS: The mean LVEF on ERNV was 47.9 ± 15.5%. The mean LVEF values for ECTb, QGS, 4D-MSPECT and Myometrix were 51.5 ± 19.6, 51.0 ± 18.6, 57.1 ± 19.3 and 49.7 ± 19%, respectively. On correlation analysis, a very strong positive correlation was observed between LVEF values derived by ERNV and those derived by the MPS software programs: ECTb (r=0.842, P<0.0001), QGS (r=0.835, P<0.0001), 4D-MSPECT (r=0.830, P<0.0001) and Myometrix (r=0.875, P<0.0001). Significant correlation was also seen for LVEFs among the four software programs. Normal cutoff values for ejection fraction on ECTb, QGS, 4D-MSPECT and Myometrix were 56, 52, 54 and 51%, respectively, using a 50% or more cutoff value on ERNV.

CONCLUSION: A strong correlation was observed among ECTb, QGS, 4D-MSPECT and Myometrix software programs when compared with ERNV and also between them for assessment of LVEF. However, there are subtle differences in the objective values of ejection fraction generated by individual software, which must be taken into account for clinical studies.