Real-Time 3D Acquisition Methods

Real-Time 3D Acquisition Methods

Several studies have demonstrated that 3D reconstruction

from serial 2D images provides accurate

anatomic information suitable for quantitative

analysis.4-8,10-15 However, this methodology is subject

to technical limitations during image acquisition

and requires significant offline data processing.

The development of RT3D echocardiographic

systems circumvents many of the disadvantages of

reconstructive methods. RT3D echocardiography

uses a transducer with ultrasound elements arranged

in a grid fashion (Figure 2). The earliest

devices, developed by von Ramm and colleagues,

3,16,17 used a sparse-array matrix transducer

transmitting at a frequency of 2.5 or 3.5

MHz. These transducers consisted of 256 nonsimultaneous

firing elements and acquired a pyramidal

volume data set measuring 60° 60° within

a single heartbeat. However, the resolution and

image quality of this first-generation sparse-array transducer

were relatively poor and often inferior to standard

2D images; frame rates were low; and the pyramidal

volume had a narrow sector angle of 60°,

resulting in an inability to accommodate larger ventricles.

Moreover, the images obtained with this system

were not volume-rendered online; instead, they consisted

of computer-generated 2D cut planes derived

from the 3D volume data set. These features limited

clinical application of this pioneering technology.

Current RT3D systems use matrix-array transducer

technology with a greater number of imaging

elements, typically containing more than 3000

imaging elements, compared with the 256 in the

sparse-array transducer. These current matrix-array

transducers offer improved resolution and are

rapidly becoming the primary technique for 3D

data acquisition in clinical and research practice.

However, recent improvements in transducer

technology have resulted in (1) a smaller transducer

footprint, (2) improved side-lobe suppression,

(3) greater sensitivity and penetration, and

(4) harmonic capabilities that may be used for

both gray-scale and contrast imaging. In addition,

these matrix-array transducers display either online

3D volume-rendered images or 2 to 3 simultaneous

orthogonal 2D imaging planes (ie, biplane

or triplane imaging).

RT3D systems generally have 3 acquisition

modes: real time (narrow), zoom (magnified), and

wide angle. The real-time mode displays a pyramidal

data set of approximately 50° 30° (Figure

3A, video clip 1).18 The zoom mode displays a

smaller, magnified pyramidal data set of 30° 30°

at a higher resolution (Figure 3B). The wide-angle

mode provides a pyramidal data set of approximately

90° 90°, which allows inclusion of a

larger cardiac volume (Figure 3C, video clip 2).

This wide-angle mode requires ECG gating, because

the wide-angle data set is compiled by

merging 4 narrower pyramidal scans obtained

over 4 consecutive heartbeats. To minimize reconstruction

artifacts, data should be acquired during

suspended respiration if possible. Although wideangle

data sets provide a larger pyramidal scan,

this is at the cost of lower resolution, which is

decreased compared with the narrow-angle 3D

mode.

Once a 3D data set is acquired, it must be sliced or

“cropped” to visualize the cardiac structures within

the pyramid (Figure 4). Multiple cropping methods

are available, but a common method displays 2 or 3

imaging planes simultaneously (video clips 3a and

3b). Each of these imaging planes can be manipulated

separately to appropriately align the cardiac

structures. Another cropping method involves a

single-slice plane that can be manually adjusted to

expose and display the cardiac structures of interest