A New Approach to Quantification of End-Diastolic Volume and Ejection Fraction from SPECT Equilibrium Radionuclide Angiocardiography: Methodology and Phantom Validation

Abstract

Conventional two-dimensional (2-D) methods for the assessments of left ventricular (LV) end-diastolic volume (EDV) and ejection fraction (EF) from planar equilibrium radionuclide angiocardiography (ERNA) cause estimation errors due to the suboptimal LV models used in the EDV and EF quantifications. This study was to develop and validate a new method for EDV and EF quantifications from single photon emission computerized tomographic (SPECT) ERNA. Six cardiac phantoms were imaged using two different gamma cameras. Planar images were acquired using a conventional camera and SPECT images were acquired and reconstructed with attenuation correction (AC) and without AC (NC) using a cardiac dedicated SPECT/CT system. SPECT-reprojected images were generated and integrated into an image sequence to create pseudo ERNA images and were quantified using two conventional and a new Yale ERNA quantification methods. Phantom EDV and EF quantified from 2-D planar ERNA were underestimated compared to the true EDV and EF values. SPECT-quantified EDVs using the conventional and Yale methods were significantly different. The EDV was markedly underestimated using the conventional methods, regardless of NC or AC. Difference in SPECT-quantified EFs between AC and NC was small and EF estimation errors for AC and NC were similar. The coefficients of variation for SPECT-quantified EDVs and EFs with AC and NC were identical. The Yale method significantly improved the accuracy of EDV quantification from the SPECT ERNA simulated in this phantom study. The precision of EF quantification using the Yale method was excellent, albeit the accuracy was fair. The Yale method for SPECT ERNA quantification may be preferred for accurate EDV and precise EF assessments.