Evaluation of biventricular function by cadmium-zinc-telluride SPECT gated tomographic radionuclide angiography: Comparison to conventional SPECT.

We compared and analyzed the consistency and repeatability of left and right ventricular ((LV/RV) functions obtained by gated-equilibrium radionuclide ventriculography (ERNV) with cadmium-zinc-telluride single-photon emission computed tomography (CZT-SPECT) and conventional SPECT (C-SPECT) with sodium iodide crystal detectors. Seventy-seven patients were included in the retrospective study. Both C-SPECT and CZT-SPECT imaging were performed on the same day. Correlations and differences in LV/RV ejection fraction (LVEF and RVEF), peak ejection rate (PER), and peak filling rate (PFR) were compared between the 2 models. Cardiac magnetic resonance (CMR) was partially used as the gold standard, and ultrasound results were included for comparative analysis. Interobserver reproducibility of each parameter obtained by the 2 cameras was compared. Between the 2 cameras, there were no significant difference in LVEF, LVPER, LVPFR, and RVPER (P > .05) and there were in RVEF and RVPFR (P < .05 or .001). The correlations (R value) were 0.831 (LVEF, excellent), 0.619 (RVEF, good), 0.672 (LVPER, good), 0.700 (LVPFR, good), 0.463 (RVPER, normal), and 0.253 (RVPFR, poor). There were no significant difference between CMR and CZT-SPECT in LVEF (P > .05) while there were between CMR and both C-SPECT and ultrasound (P < .05). The correlations were all good (R = 0.660, 0.658, and 0.695). There were no significant difference between CMR and both C-SPECT and CZT-SPET in RVEF (P > .05) and the correlations were good (R = 0.771 and 0.745). For repeatability, the intraclass correlation coefficient of RVPFR by C-SPECT was good (intraclass correlation coefficient = 0.698) and excellent for the rest of the groups (0.823-0.989). The repeatability of LVEF and RVEF was better for CZT-SPECT than for C-SPECT. The repeatability of PER was better for both cameras than PFR. CZT-SPECT tomographic ERNV correlated well with C-SPECT planar ERNV in evaluation of biventricular systolic function and LV diastolic function. Compared with the "gold standard" CMR, both models had good correlation in measuring LV/RVEF. CZT-SPECT had better inter-group reproducibility than C-SPECT. The accuracy of RV diastolic function need further study. CZT-SPECT tomographic ERNV will play an important and unique role in the clinical application of accurate evaluation of biventricular function in the future.