Assessment of left ventricular function with computed tomography and magnetic resonance in patients with low ejection fraction scheduled for coronary artery bypass grafting: a preliminary study

INTRODUCTION Reliable estimation of left ventricular (LV) parameters with the use of computed tomography (CT) has not been verified in patients with low ejection fraction (EF) so far.
OBJECTIVES The aim of this preliminary study was to assess the agreement between magnetic resonance imaging (MRI) and CT in the assessment of the LV volume and EF in patients with low EF scheduled for coronary artery bypass grafting (CABG). 
PATIENTS AND METHODS In 18 patients (2 women, 16 men) with the EF of 40% or lower scheduled for CABG, cardiac 1.5T MRI and 16‑slice CT were performed within 48 hours before surgery. All MRI scans were analyzable; 5 CT exams were excluded due to poor quality. Between‑method agreement was assessed using the analysis of correlation and the Bland‑Altman plots.
RESULTS The end‑systolic volume (ESV), end‑diastolic volume (EDV), and stroke volume (SV) were higher when measured by MRI compared with CT (P <0.05 for all). The EF values measured by both methods were comparable. We observed a strong positive correlation between MRI and CT in the measurement of ESV (R = 0.86), EDV (R = 0.71), and EF (R = 0.68), but there was no correlation for SV (R = 0.07). The Bland‑Altman analysis confirmed that the LV volumes assessed by MRI were higher compared with those determined by CT. There was also a trend for larger differences between the 2 methods in the measurement of ESV and EDV. The EF value was higher on MRI compared with CT but the difference was nonsignifiact.
CONCLUSIONS In patients with coronary artery disease and low EF, the EF values measured by MRI and CT are comparable, but the ESV, EDV, and SV parameters cannot be used interchangeably due to their underestimation on CT.