Quantification of left ventricular parameters obtained by automated software for 64-slice multidetector computed tomography and comparison with magnetic resonance imaging

Abstract

The dataset obtained with 64-slice multidetector CT (MDCT) for coronary artery evaluation can be used to calculate important left ventricular (LV) volumetric parameters. We compared LV parameters derived by new, commercially available, fully automated software for MDCT (Syngo Circulation, Siemens, Germany) to cardiac magnetic resonance (CMR) as a reference standard. Twenty patients underwent CMR after completing a clinically indicated MDCT. Ejection fraction (EF), end-systolic volume (ESV), end-diastolic volume (EDV), stroke volume (SV), and myocardial mass (MM) for MDCT were obtained using automated software and were compared to CMR measurements, with papillary muscles (PMs) included in, or excluded from, the blood pool. The Pearson correlation coefficient (r) and Bland-Altman method were used to determine agreement between methods. When PMs were included in the blood pool, the correlation was excellent for EF (r=0.92, p\0.001), ESV (r=0.86, p\0.001), and EDV (r=0.80, p\0.001). When PMs were excluded from CMR, correlation was still very good for EF (r=0.89, p\0.001), ESV (r=0.82, p\0.001), and EDV (r=0.82, p\0.001). MDCT values for SV and MM showed a good correlation compared to both CMR methods. When PMs were included, the correlation was good for SV (r=0.70, p\0.001), and MM (r=0.70, p\0.001); when they were excluded, the correlation was less robust but still significant for SV (r=0.71, p\0.001) and MM (r=0.73, p\0.001). In conclusion, EF, ESV, and EDV obtained by MDCT using simple, automated software correlated very well with CMR; SV and MM showed good correlation. Automated analysis of volumetric parameters by MDCT can be reliably utilized for clinical purposes.