The initial absolute measurement and subsequent monitoring of left ventricular volume is an important clinical parameter in patients with cardiac disease. Although cineventriculography has been accepted as a clinical standard for left ventricular volume determina- tion, the invasive nature of the technique and the inherent assumptions concerning the geometry of the left ventricle limit its application in repeated assess- ments. 1-4 Radionuclide methods offer an alternative noninvasive approach 5 but are subject to other limita- tions. These are mainly related to the detection of edges and end planes as well as the determination of the level of background activity. 6 Previous attempts with standard two-dimensional echocardiographic tech- niques have also demonstrated several important limi- tations, particularly in patients with regions of left ven- tricular asynergy. Factors contributing to the low predictive accuracy of two-dimensional echocardio- graphic volume measurements include geometric assumptions, image plane positioning errors, and Three-dimensional echocardiographic evaluation of left ventricular volume: Comparison of Doppler myocardial imaging and standard gray-scale imaging with cineventriculography—an in vitro an in vivo study Aleksandra Lange, MD, Przemystaw Palka, MD, Andrzej Nowicki, PhD, Robert Olszewski, MD, Thomas Anderson, MSc, Jerzy Adamus, MD, George R. Sutherland, MD, and Keith A.A. Fox, MD Edinburgh, United Kingdom, and Warsaw, Poland Background Standard gray-scale imaging (GSI), three-dimensional (3D) echocardiography has been shown to superior to two-dimensional echocardiography in measuring left ventricular volume. However, the often relative ity of transthoracic gray-scale data can limit the potential application of this technique. Doppler myocardial imag a new ultrasound technique that potentially offers higher-quality 3D images with a transthoracic approach than technique. This study was designed to compare the accuracy of standard GSI and DMI 3D left ventricular volume ments in vitro and in vivo. Methods and Results In vitro, the minimum and maximum volume of the contracting single-chamber, tissue king phantom was calculated by using both techniques. In vivo, GSI and DMI 3D left ventricular volume measure performed in 16 patients. End-diastolic and end-systolic left ventricular volumes were computed for both techniq pared with those calculated by cineventriculography. In vitro, both methods tended to underestimate the true p ume, but the systematic error was smaller for DMI than for GSI (–1.2% ± 1.5% vs. –4.3% ± 3%; p < 0.01) and was more constant in the case of DMI over the range of different sizes of true volume. In vivo, for GSI the end-diastolic volu difference was –12.6 ml and the limits of agreement were ±18 ml, and for DMI the corresponding values were –4 ±10.6 ml, respectively. The difference for end-systole was –6.5 ± 10.6 ml and –1.5 ± 10 ml for GSI and DMI, resp The magnitude of the difference in volume measurement between 3D echocardiography and cineventriculograp nificantly smaller when using the Doppler technique. Conclusions The results of this in vitro and in vivo study indicate that DMI is superior to GSI as a transthoracic tion technique for 3D volume computation. (Am Heart J 1998;135:970-9.) From the Department of Cardiology, Western General Hospital, Edinburgh; Polish Academy of Sciences, Warsaw; Department of Cardiology, Central Clinical Hospi- tal, Military Medical School, Warsaw; the Department of Medical Physics and Med- ical Engineering, Royal Infirmary, Edinburgh; and the Cardiovascular Research Unit, University of Edinburgh. Supported by a grant from the Scottish Office Home and Health Department (No. K/MRS/50/C2357). Submitted April 29, 1997; accepted Dec. 18, 1997. Reprint requests: Dr. A. Lange, Department of Cardiology, Western General Hospi- tal, Crewe Road, Edinburgh EH4 2XU, United Kingdom. Copyright © 1998 by Mosby, Inc. 0002-8703/98/$5.00 + 04/1/88877