Validation of 4D-MSPECT and QGS for quantification of left ventricular volumes and ejection fraction from gated 99mTc-MIBI SPET:: comparison with cardiac magnetic resonance imaging

The main aim of this study was to validate the accuracy of 4D-MSPECT in the assessment of left ventricular (LV) end-diastolic/end-systolic volumes (EDV, ESV) and ejection fraction (LVEF) from gated technetium-99m methoxyisobutylisonitrile single-photon emission tomography (Tc-99m-MIBI SPET), using cardiac magnetic resonance imaging (cMRI) as the reference method. By further comparing 4D-MSPECT and QGS with cMRI, the software-specific characteristics were analysed to elucidate clinical applicability. Fifty-four patients with suspected or proven coronary artery disease (CAD) were examined with gated Tc-99m-MIBI SPET (8 gates/cardiac cycle) about 60 min after tracer injection at rest. LV EDV, ESV and LVEF were calculated from gated Tc-99m-MIBI SPET using 4D-MSPECT and QGS. On the same day, cMRI (20 gates/cardiac cycle) was performed, with LV EDV, ESV and LVEF calculated using Simpson's rule. Both algorithms worked with all data sets. Correlation between the results of gated Tc-99m-MIBI SPET and cMRI was high for EDV [R=0.89 (4D-MSPECT), R=0.92 (QGS)], ESV [R=0.96 (4D-MSPECT), R=0.96 (QGS)] and LVEF [R=0.89 (4D-MSPECT), R=0.90 (QGS)]. In contrast to ESV, EDV was significantly underestimated by 4D-MSPECT and QGS compared to cMRI [130+/-45 ml (4D-MSPECT), 122+/-41 ml (QGS), 139+/-36 ml (cMRI)]. For LVEF, 4D-MSPECT and cMRI revealed no significant differences, whereas QGS yielded significantly lower values than cMRI [57.5%+/-13.7% (4D-MSPECT), 52.2%+/-12.4% (QGS), 60.0%+/-15.8% (cMRI)]. In conclusion, agreement between gated Tc-99m-MIBI SPET and cMRI is good across a wide range of clinically relevant LV volume and LVEF values assessed by 4D-MSPECT and QGS. However, algorithm-varying underestimation of LVEF should be accounted for in the clinical context and limits interchangeable use of software.