The Transaxial Orientation Is Superior to Both the Short Axis and Horizontal Long Axis Orientations for Determining...

We sought to determine which of the three orientations is the most reliable and accurate for quantifying right ventricular (RV) volume and ejection fraction (EF) by cardiac magnetic resonance using Simpson's method. We studied 20 patients using short axis (SA), transaxial (TA), and horizontal long axis (HLA) orientations. Three readers independently traced RV endocardial contours at end-diastole and end-systole for each orientation. End-diastolic volumes (EDVs), end-systolic volumes (ESVs), and EF were calculated and compared with the 3D piecewise smooth subdivision surface (PSSS) method. The intraclass correlation coefficients among the 3 readers for EDV, ESV, and EF were 0.92, 0.82, and 0.42, respectively, for SA, 0.95, 0.92, and 0.67 for TA, and 0.85, 0.93, and 0.69 for HLA. For mean data there was no significant difference between TA and PSSS for EDV (-2.6%, 95% CI: -8.2 to 3.3%), ESV (-5.9%, -15.2 to 4.5%), and EF (1.7%, -1.5 to 4.9%). HLA was accurate for ESV (-8.9%, -18.5 to 1.8%) and EF (-0.7%, -3.8 to 2.5%) but significantly underestimated EDV (-9.8, -16.6 to -2.4%). SA was accurate for EDV (0.5%, -6.0 to 7.5%) but overestimated ESV (10.5%, 0.1 to 21.9%) and had poor interrater reliability for EF. Conclusions. The TA orientation provides the most reliable and accurate measures of EDV, ESV, and EF.