Optimal number of image slices for reliable measurement of left ventricular volumes and ejection fraction by three-dimensional transesophageal echocardiography

Purpose Quantification of the left ventricular (LV) volume by three-dimensional echocardiography is accurate but time-consuming. To shorten the time required, we sought to determine the minimum number of image planes necessary to measure LV volume reliably.Methods We analyzed transesophageal three-dimensional echocardiographic LV data obtained by the rotational scanning method in 16 patients: 11 had ischemic heart disease, and 5 had dilated cardiomyopathy. LV volumes were calculated from 6, 10, and 30 short-axis images using the disk-summation method and from 2, 4, 6, 10, 20, and 30 longitudinal images using the new average rotation method.Results LV volume varied less with the average rotation method than with the disk-summation method. The 95% limit of agreement between the 30-image and 6-image methods was 0.3% ± 3.7% for the average rotation method, whereas it was –2.0% ± 6.9% for the disk-summation method. The time required for analysis decreased from 12.5 ± 2.8min with the 30-image method to only 3.3 ± 0.5min for the 6-image method.Conclusions Measurement of six longitudinal images provided reliable LV volume data, even in patients with enlarged or deformed left ventricles. The short measurement time supports the use of three-dimensional echocardiographic LV volume measurement in the clinical setting.