A fast and reproducible method to estimate left atrial volume using cardiac computed tomography

The purpose of this study was to do better than the simple maximal axial area (MArea_ax) and to validate simple, fast and robust orthogonal methods for determining the left atrium volume (LAV) with cardiac CT (CCT).

A total of 60 patients who underwent CTT were retrospectively included. There were 30 men and 30 women with a mean age of 59±12 (SD) years (range: 27-80 years), using four methods to assess LAV beyond the MArea_ax: two orthogonal methods requiring the measurements of axial, coronal and sagittal planes of 3 diameters LAV_3diam or 3 area LAV_3Areas; Area-length biplane method LAV_biplane; and volumetric method LAV_volumetric using a semi-automated tool that served as the reference standard. The orthogonal methods were applied on contrast-enhanced (IV+) and unenhanced (IV-) CCT images. Comparisons were performed using Pearson correlation test (r) and Bland-Altman analysis. Inter- and intra-observer variability were assessed using intraclass correlation coefficient (ICC) with a two-way mixed-effects model.

On IV+ CCT, LAV_biplane, LAV3diam_-IV+, LAV_3Areas-IV+ underestimated LAV (-15±1.99mL; -21±1.37mL; -15±1.98mL; all P<0.001). LAV_3diam-IV+, LAV3_Areas-IV+ better correlated with reference standard (r=0.97 and 0.98) than LAV_biplane (r=0.79) as well as MArea_ax (r=0.90). Estimating LAV on IV- further showed high correlation against the reference (r=0.93 and 0.95 for LAV3diam_-IV- and LAV_3Areass-IV-, respectively). Intra- and inter-observer ICC increased from LAV_volumetric (2.43% and 3.09%); LAV_3Areas-IV+ (3.04 and 3.30%); LAV_3Areas-IV-(3.34 and 4.23%), LAV3diam_-IV+ (3.36 and 5.11%); LAV3diam_-IV- (5.16 and 6.90%); to LAV_biplane (9.65 and 10.28%).

Better than MArea_ax, orthogonal methods using either diameter or surface are fast and reproducible methods to assess LAV on CCT when performed with or without intravenous administration of contrast material.