Performance et reproductibilité de TOMPOOL pour l’estimation des fractions d’éjection, volumes et débits bi-ventriculaires

Abstract

BackgroundTo evaluate the diagnostic accuracy of TOMPOOL (gated blood pool SPECT). Comparison with a reference-algorithm: BP-SPECT. Comparison with MRI and right heart catheter measurement. Estimation of the inter-observer reproducibility. Methods and resultsNinety-nine consecutive patients underwent GBPS and two physicians performed the measurement using TOMPOOL and BP-SPECT. Results were compared to the recommended reference-standard: cardiac magnetic resonance imaging [CMR] (EDV, ESV, EF, n=35), catheter measurements from thermodilution [TD] (CO, n=21). ResultsResults are always referred using the following order: correlation with BP-SPECT (n=99), correlation with reference-standard (MRI or right heart catheter measurement), inter-observer correlation. RVEF: r=0.67; r=0.70; r=0.81. RVEDV: r=0.74; r=0.49; r=0.77. RVESV: r=0.76; r=0.64; r=0.81. RCO: r=0.68; 0.63; 0.78. LVEF: r=0.88; r=0.79; r=0.94. LVEDV: r=0.82; r=0.80; r=0.90. LVESV: r=0.89; r=0.90; 0.95. ConclusionTOMPOOL allows a quantitative and qualitative analysis of the volume and systolic function of the left ventricle. However, a prior CMR-based calibration is preferable (thresholds must be adjusted). A qualitative analysis of the right ventricle is more reliable (poorer correlation and reproducibility).