Application of the R0-R3 formulas using the ECToolbox software to calculate left ventricular ejection fraction in myocardial perfusion SPET and comparison with equilibrium radionuclide ventriculography. Normal cutoff values for a Chinese population.

The aim of this study was to compare the correlation and consistency of left ventricular ejection fraction (LVEF) obtained by ECG-gated myocardial perfusion SPET (GMPS) using the four formulas (R0-R3) in ECToolbox software and by equilibrium radionuclide ventriculography (ERNV), and determine the optimal diagnostic thresholds of the four formulas in a Chinese population. A hundred and three candidate donors (59 male and 44 female), including 38 patients with a history of myocardial infarction and 65 patients with suspected coronary heart disease, underwent both (99m)Tc-MIBI rest GMPS and technetium-99m red blood cells ((99m)Tc-RBC) ERNV within a week. The LVEF values calculated by ECToolbox R0, R1, R2 and R3 were compared with those obtained by ERNV. Using LVEF≥50% obtained by ERNV as the gold standard, the optimal diagnostic thresholds of the four formulas (R0-R3) were assessed by receiver operating characteristic (ROC) curves. Results showed that the mean LVEF value of ERNV was 54.6±17.5%, and the mean LVEF value of the four formulas was 64.1±15.7%, 56.3±15.1%, 69.9±17.9% and 56.3±13.6%, respectively, showing a significantly strong correlation between the results obtained by the two methods (r>0.85, P<0.001). All mean LVEF values obtained by the four formulas were higher than the mean LVEF value obtained by ERNV, and there was very significant difference between R0 and R2 results and the ERNV result (t=12.511 and 18.652, P<0.001). Furthermore, there was significant difference between R1 and R3 results and the ERNV result (t=2.169 and 2.570, P<0.05). Using ERNV LVEF≥50% as the normal diagnostic value, the optimal diagnostic threshold of R0∼R3 was 56.5%, 51.5%, 64.5% and 52.5%, respectively. There was a strong correlation between the LVEF values obtained by the four formulas in ECToolbox software and ERNV, but the numerical values of LVEF differed between the four formulas. In conclusion, A strong correlation was observed among R0, R1, R2 and R3 in the ECToolbox software when compared with ERNV and also between them for the assessment of LVEF. However, there were some differences in the numerical values of LVEF generated by the individual formulas, which must be taken into account in comparing clinical studies.