Nosir Y F, Salustri A, Kasprzak J D, Breburda C S, Ten Cate F J, Roelandt J R
Heart Centre and Thoraxcenter, Division of Cardiology, University Hospital Rotterdam-Dijkzigt, Rotterdam, The Netherlands.
J Am Soc Echocardiogr. 1998 Jun;11(6):620-30. doi: 10.1016/s0894-7317(98)70038-0.
Serial evaluation of left ventricular (LV) ejection fraction (EF) is important for the management and follow-up of cardiac patients. Our aim was to compare LVEF calculated from two three-dimensional echocardiographic (3DE) methods with multigated radionuclide angiography (RNA), in patients with normal and abnormally shaped ventricles.
Forty-one consecutive patients referred for RNA underwent precordial rotational 3DE acquisition of 90 cut-planes. From the volumetric data set, LVEF was calculated by (a) Simpson's rule (3DS) through manual endocardial tracing of LV short-axis series at 3 mm slice distance and (b) apical biplane modified Simpson's method ( MS) in 29 patients by manual endocardial tracing of the apical four-chamber view and its computer-derived orthogonal view. Patients included three groups: A, 17 patients with LV segmental wall motion abnormalities; B, 13 patients with LV global hypokinesis; and C, 11 patients with normal LV wall motion. For all the 41 patients, there was excellent correlation, close limits of agreement, and nonsignificant difference between 3DS and RNA for LVEF calculation (r = 0.99, [-6.7, +6.9] and p = 0.9), respectively. For the 29 patients, excellent correlation and nonsignificant differences between LVEF calculated by both 3DS and BMS and values obtained by RNA were found (r = 0.99 and 0.97, p = 0.7 and p = 0.5, respectively). In addition, no significant difference existed between values of LVEF obtained from RNA, 3DS, and BMS by the analysis of variance (p = 0.6). The limits of agreement tended to be closer between 3DS and RNA (-6.8, +7.2) than between BMS and RNA (-8.3, +9.7). The intraobserver and inter-observer variability of RNA, 3DS, and BMS for calculating LVEF(%) were (0.8, 1.5), (1.3, 1.8), and (1.6, 2.6), respectively. There were closer limits of agreement between 3DS and RNA for LVEF calculation in A, B, and C patient subgroups [(-3.5, +5), (-8.4, +5.6), and (-7.8, +8.6)] than that between BMS and RNA [(-8.1, +10.7), (-11.9, +9.3), and (-9.1, +11.3)], respectively.
No significant difference existed between RNA, 3DS, and BMS for LVEF calculation. 3DS has better correlation and closer limits of agreement than BMS with RNA for LVEF calculation, particularly in patients with segmental wall motion abnormalities and global hypokinesis. 3DS has a comparable observer variability with RNA. Therefore the use of 3DS for serial accurate LVEF calculation in cardiac patients is recommended.
对左心室（LV）射血分数（EF）进行连续评估对于心脏病患者的管理和随访至关重要。我们的目的是比较在心室形状正常和异常的患者中，两种三维超声心动图（3DE）方法计算的左心室射血分数与多门控放射性核素血管造影（RNA）的结果。
41例连续接受RNA检查的患者接受了心前区90个切面的旋转3DE采集。从容积数据集中，通过以下方法计算左心室射血分数：（a）辛普森法则（3DS），通过在3mm切片距离处手动追踪左心室短轴系列的心内膜；（b）对29例患者采用心尖双平面改良辛普森法（MS），通过手动追踪心尖四腔心切面及其计算机衍生的正交切面的心内膜。患者分为三组：A组，17例左心室节段性室壁运动异常患者；B组，13例左心室整体运动减弱患者；C组，11例左心室壁运动正常患者。对于所有41例患者，在计算左心室射血分数时，3DS与RNA之间具有极好的相关性、相近的一致性界限且差异无统计学意义（r = 0.99，[-6.7, +6.9]，p = 0.9）。对于29例患者，发现3DS和BMS计算的左心室射血分数与RNA获得的值之间具有极好的相关性且差异无统计学意义（r分别为0.99和0.97，p分别为0.7和0.5）。此外，通过方差分析，RNA、3DS和BMS获得的左心室射血分数值之间无显著差异（p = 0.6）。3DS与RNA之间的一致性界限（-6.8, +7.2）比BMS与RNA之间的一致性界限（-8.3, +9.7）更接近。RNA、3DS和BMS计算左心室射血分数（%）的观察者内和观察者间变异性分别为（0.8, 1.5）、（1.3, 1.8）和（1.6, 2.6）。在A、B和C患者亚组中，3DS与RNA计算左心室射血分数的一致性界限[（-3.5, +5）、（-8.4, +5.6）和（-7.8, +8.6）]比BMS与RNA之间的一致性界限[（-8.1, +10.7）、（-11.9, +9.3）和（-9.1, +1 .3）]更接近。
在计算左心室射血分数方面，RNA、3DS和BMS之间无显著差异。在计算左心室射血分数时，3DS与RNA的相关性优于BMS且一致性界限更接近，尤其是在节段性室壁运动异常和整体运动减弱的患者中。3DS与RNA的观察者变异性相当。因此，建议在心脏病患者中使用3DS进行连续准确的左心室射血分数计算。
