Quantitative evaluation of left atrial volumes and ejection fraction by 320-slice computed-tomography in comparison with three- and two-dimensional echocardiography: a single-center retrospective-study in 22 subjects.

PURPOSE

To evaluate efficacy and reproducibility of 320-slice computed tomography (CT) for measuring left atrial (LA) maximum (LAVmax) and minimum volume (LAVmin) during the cardiac cycle, we compared CT with three- and two-dimensional (3D and 2D) transthoracic echocardiogram (TTE).

MATERIALS AND METHODS

LAVmax and LAVmin (ml), and LA ejection fraction (LAEF) (%) were assessed in 22 consecutive subjects (15 males, 59.5 ± 15.1 years) using retrospective electrocardiogram gated 320-slice CT (Aquilion One, Toshiba Medical) and 3D-TTE (IE-33, Phillips). LAVmax and LAVmin were selected from the time volume curve. LAEF was calculated as (LAVmax-LAVmin)/LAVmax×100(%).

RESULTS

Mean ± standard deviation (SD) of LAVmax and LAV min were significantly larger by CT than 3D-TTE or 2D-TTE (both P<0.01). LAEF was 25.3 ± 13.1% by CT, 30.2 ± 6.8% by 3D-TTE (P=NS) and 33.9 ± 8.9% by 2D-TTE (P<0.05). The correlation coefficients (CCs) between CT and 3D-TTE in LAVmax, LAmin, and LAEF were 0.64, 0.68 and 0.57, respectively. Mean difference ± 1.96SD of LAVmax, LAVmin and LAEF by Bland and Altman analysis calculated from CT minus 3D-TTE were 48.8 ± 59.1 ml, 41.1 ± 63.2 ml, and -4.9 ± 21.1%, respectively. The CCs and mean difference ± 1.96SD between CT and 2D-TTE had similar tendencies. The CCs of interobserver variation were (for CT, 3D-TTE, 2D-TTE, respectively): 0.90, 0.95 and 0.94 (LAVmax), 0.97 and 0.97 and 0.93 (LAVmin), and 0.64, 0.77 and 0.34 (LAEF).

CONCLUSIONS

320-slice CT enables direct LAV measurements and has high reproducibility and positive correlation with 3D and 2D TTE. Absolute value of LAV by CT was larger than that by 3D and 2D TTE.