Left ventricular ejection fraction by real-time three-dimensional echocardiography : The Necker cube for the naive realism of two-dimensional methods.

Abstract

Despite a staggering technical progress in echocardiography, in daily clinical practice assessment of left ventricular (LV) ejection fraction (EF) is still often done in a rather inaccurate two-dimensional (2D) manner. Real-time three-dimensional echocardiography (RT3DE) provides a more representative realistic view of LV volumes, thereby being the Necker cube for the naive realism of 2D methods. The Necker cube (Fig. 1a) is an optical illusion first published by the Swiss crystallographer Louis Albert Necker in 1832 [1]. It provides a counter-attack against naive realism, which states that the way we observe the world is the way the world actually is. The Necker cube contradicts this claim, because we see one or the other of two cubes (Fig. 1b), but really, there is no cube there at all, but only a 2D drawing of 12 lines. We see something which is not really there, thus apparently disproving naive realism. This criticism of naive realism supports representative realism. This view argues that we experience reality indirectly by perceptions that represent the real world. For example, if we see a yellow flower, we do not actually see the flower itself but a representation of it. In this way, differences of perception which occur due to changes in the position of the viewer, light conditions, and so on can be easily explained: it is not the object that is changing, only our perception of it. Fig. 1 a The Necker cube. b Like a 2D echocardiographic image, the Necker cube may give the suggestion of a realistic representation of a 3D structure, but really there is only a 2D picture that can be interpreted in different ways Let us go back to echocardiography. Visual estimation of wall motion (‘eyeballing’) is highly subjective, imprecise and poorly reproducible. Nevertheless, some sense of naive realism seems to provide an unrealistic confidence in the accuracy of our eyes. Calculating LV volumes and EF from 2D images by using algorithms such as the biplane method of disks (modified Simpson’s rule) are considered more accurate and better reproducible. However, because of the geometric assumptions of this 2D method, measurements may be inaccurate if the shape of the left ventricle is abnormal or when the acquisition of the 2D images is suboptimal. In other words, like the yellow flower, 2D echocardiographic images are not necessarily a realistic representation of the true left ventricle, but only a perception that is influenced by subjective interpretation and variability in measurements of only a very small 2D portion of the true 3D volume.