Interstudy repeatability of left and right ventricular volume estimations by serial-gated tomographic radionuclide...

Estimation of left ventricular ejection fraction (LVEF) with 99MTc-HSA equilibrium radionuclide angiography (RA) is frequently used for assessing cardiac function. The purpose of this study was to investigate the interstudy repeatability of left (LV) and right (RV) ventricular volume and ejection fraction estimations, using a cadmium-zinc-telluride (CZT) SPECT camera.

Forty-six patients were scanned twice, interrupted by repositioning. Each acquisition was analysed twice by two experienced technologists. Interstudy and interobserver variations were calculated as the coefficient of variation (CV) and the 95% confidence interval (CI) for limits of agreement (LOA) between each sequence of analyses for each of the two acquisitions.

The Interstudy variations and 95% CI for LVEF, LV end systolic (LVESV) and end diastolic (LVEDV) volumes were 4.6% (-5.4 to 6.4), 9.3% (-6.90 to 5.20) and 7.0% (-13.9 to 11.1), respectively. For the right ventricle, the corresponding values were 11.9% (-9.40 to 10.8), 9.8% (-14.9 to 10.8) and 8.1% (-20.7 to 16.3).

The CZT detector camera has excellent reproducibility with regard to interstudy variation when assessing LV volumes and EF. Interstudy variation is considerably higher for RV volumes and EF, indicating a lack of consistency in tracing the boarders of the right ventricle.