Measurement of left ventricular ejection fraction and volumes with three-dimensional reconstructed transesophageal ultrasound scans: comparison to radionuclide and thermal dilution measurements.

A transesophageal, ultrasonic cardiac imaging probe was built that incorporated a mechanism for changing the angle of the imaging plane of a conventional phased array in a precise and known manner. This probe was used to acquire an angular spatial sequence of two-dimensional images of the left ventricular cavity over a series of cardiac cycles by sweeping the imaging plane through it stepwise. The endocardial borders of these images were manually outlined off-line and the application of a three-dimensional reconstruction algorithm was then used to compute the left ventricular end-diastolic and end-systolic volumes and ejection fraction. A study was conducted with seven anesthesized dogs to compare ultrasonic determinations by this method with determinations and measurements made using radionuclide and thermal dilution methods. Comparison of 33 ejection fractions, measured by the ultrasonic volume method and by the gated blood pool radionuclide approach, yielded a correlation coefficient of 0.87 and a standard error of the estimate of 5.7% measured over a range of 10% to 58% (average, 40%). Comparison of the ultrasonically measured volumes with those calculated from stroke volume (derived from thermal dilution cardiac output measurement) and ejection fraction (measured by radionuclide technique) produced a correlation coefficient of 0.92 and a standard error of the estimate of 10.3 mL over a range of 18 to 130 mL (average, 56 mL). The accuracy of volume and ejection fraction measurements with this new ultrasonic method seems comparable to that of other currently used clinical approaches such as radionuclide and angiography.