A Survey of Three-Dimensional Modeling Techniques for Quantitative Functional Analysis of Cardiac Images - Advance Image Processing in Magnetic Resonance Imaging - page 261

Left
auricle

Aorta

Superior vena cava

Pulmonary
artery

Right auricle

Tricuspid
valve

Right
ventricle

Pulmonary
veins
Left
ventricle
Mitral valve

Inferior vena cava

Pulmonary
valve

FIGURE 9.1 The heart is a complex system whose aim is to pump blood to irrigate the

whole body. Assessment of cardiac performance is crucial in diagnosis and treatment of
cardiovascular diseases.

interobserver variability. Therefore, many researchers have addressed the problem
of automatic left (LV) and right (RV) ventricle segmentation. Because the shape of
the cardiac ventricles is approximately known, it seems natural to incorporate this
knowledge into the segmentation process. Such model-driven techniques have
received ample attention in medical image analysis in the last decade [3,4]. Some
of the advantages over model-free approaches are: (1) the model itself can constrain
the segmentation process that is ill posed in nature owing to noise and image
artifacts, (2) segmentation, image analysis, and shape modeling are simultaneously

addressed in a common framework, (3) models can be coarse or detailed depending

on the desired degree of abstraction, and (4) in some approaches, most of the
chamber's shape can be explained with a few comprehensible parameters that can

subsequently be used as cardiac indices (cf. References [ 5 - 9 ] among others).

Use of geometric models is not completely new to the analysis of cardiac
images. As a matter of fact, traditional methods of obtaining parameters such as
left-ventricular volume (LVV) and mass (LVM) from echo- and angiocardio-
graphy were based on (simple) geometrical models [10-13]. However, their use
was mainly motivated by the need to extract 3-D parameters from 2-D images,
and their accuracy was therefore limited [14].
The literature on model-driven segmentation of cardiac images has grown
rapidly in the last few years, and this trend is likely to continue.

*

This chapter

See, for instance, a recently published special issue on 3-D cardiac image analysis [ 15 ].

*