It is generally believed that three-dimensional (3D) echocardiography can provide unique en face views of atrial septal defects (ASDs), whereas conventional two-dimensional (2D) transthoracic echocardiography (TTE) cannot. The purpose of this study was to develop a special 2D TTE–based en face view of ASDs.
En face views of ASDs on 2D TTE were obtained in 415 consecutive adult patients and compared with the results of real-time 3D transesophageal echocardiography in 25 of these patients.
Acceptable en face views of ASDs on 2D TTE were obtained in 80% of patients, in whom secundum ASDs could be adequately imaged in more standard 2D imaging planes. The ability to visualize en face views of ASDs on 2D TTE was inversely related to their sizes ( P < .05). The differences in the major and minor dimensions of ASDs between 2D TTE and real-time 3D transesophageal echocardiography were not statistically significant ( P > .05). Conventional 2D transthoracic echocardiographic views significantly underestimated ASD size with superior-inferior directional major axis.
Two-dimensional TTE can provide en face views of ASDs and their spatial orientations with respect to neighboring structures. This method may provide incremental information to 3D echocardiography in patients with ASDs.
Real-time three-dimensional (RT3D) transesophageal echocardiography (TEE) is a recently developed technique that is increasingly used in echocardiography laboratories. Generally, atrial septal defects (ASDs) have complex shapes that are not well characterized on two-dimensional (2D) transthoracic echocardiography (TTE), whereas three-dimensional (3D) echocardiography provides unique en face views of ASDs. However, unlike conventional 2D transthoracic echocardiographic views, the special 2D transthoracic echocardiographic imaging technique may provide en face views of ASDs and their spatial orientations with respect to neighboring structures. This study was designed to evaluate the feasibility and advantages of en face views of secundum ASDs obtained using 2D TTE and compared with findings on RT3D TEE as a reference.
In this prospective study, 415 consecutive adult patients (120 male, 295 female; age range, 14-75 years; mean age, 35 years) with ostium secundum ASDs and significant left-to-right shunts were referred to our institution from August 2006 to September 2009. Patients with poor transthoracic echocardiographic images and ≥2 ASDs were excluded. All patients had adequate 2D transthoracic echocardiographic images for evaluating ASDs. After we acquired RT3D transesophageal echocardiographic capabilities in May 2009, 25 patients (8 men, 17 women; age range, 20-61 years; mean age, 40 years) were also evaluated using RT3D TEE as well as 2D TTE. Patients were stratified by New York Heart Association functional classification. Written informed consent was obtained from all patients. The study was approved by our institutional review board.
En Face Views of ASDs on 2D TTE
Patients with ASDs were diagnosed using TTE. Echocardiographic images were obtained in all subjects in a steep left lateral decubitus position, with the patient’s left arm extended over the head, by an experienced sonographer using an Acuson Sequoia C256 (Siemens Medical Solutions, Mountain View, CA) echocardiographic imaging system equipped with a 2.0-MHz to 3.5-MHz transducer. An electrocardiogram was recorded and displayed simultaneously on the image. There were two methods to image en face views of ASDs on 2D TTE.
First, the ASD was shown in the parasternal or apical 4-chamber view. Because all patients had enlarged right ventricles, the apical area was essentially showing the right ventricle, and the view became a nonstandard 4-chamber view. We adjusted the transducer position with slight movement and rotation and angulation to make the ultrasonic beam parallel to the interatrial septum ( Figure 1 A). The image index marker was at approximately 1 o’clock. The majority of patients did not require repositioning. Keeping the atrial septum and ASD in the region of interest, the transducer was rotated counterclockwise approximately 45° to 60°, such that the direction of the ultrasound beam remained similar to that of the aforementioned nonstandard 4-chamber view, and the image index marker was at approximately 11 o’clock. The interatrial septal plane and en face view of the ASD could be obtained ( Figure 1 B). Minor adjustment may be required to achieve the view, because the region of interest may depart from the interatrial septal plane during transducer rotation.
Second, the transducer was placed in the parasternal and apical area. A 4-vessel (ie, pulmonary artery, aorta, and superior and inferior vena cava; Figure 2 A) or 3-vessel (ie, pulmonary artery, aorta, and superior vena cava) view of the heart can be obtained by rotating the transducer approximately 45° to 60° counterclockwise from the apical 4-chamber view with 20° angulation toward the right side of patient and slight tilting anteriorly. The ultrasound beam was directed approximately toward the patient’s right shoulder, and the image index marker was at approximately 11 o’clock. If the transducer was slightly angulated toward the patient’s left lateral side from the 4-vessel or 3-vessel view, the en face view of the ASD and the interatrial septal plane could also be shown ( Figure 2 B). When patient had an enlarged right ventricle, it was easier to show the 4-vessel or 3-vessel view of the heart. To avoid instrument artifacts and echo dropout, en face views of ASDs must be confirmed using color flow mapping ( Figure 2 C). We used a dynamic range of approximately 70 dB and tissue harmonic wave mode. With respect to the optimal image, slight tilting or angulation of the transducer may be necessary. Because the atrial septum may not be a plane, and also because there are individual variations in the position of the heart within the thorax, the transducer position may vary among subjects. All echocardiographic data were stored digitally on magneto-optical discs for online and subsequent analysis.
The image quality of TTE depicting en face views of ASDs was evaluated subjectively by two independent experienced cardiac echocardiographers offline using a 4-point grading scale: excellent, good, moderate, or poor. An excellent image was an en face view of an ASD with clear continuous border visualization ( Figure 3 A). A good image showed a distinct border of an en face view of an ASD with >75% visualization ( Figure 3 B). A moderate-quality image was an en face view of an ASD with 50% to 75% visualization ( Figure 3 C). A poor image was an en face view of an ASD with adequate border delineation of <50%. Images of moderate and poor quality were considered unacceptable for en face views of ASDs on 2D TTE.
En Face Views of ASDs on RT3D TEE
According to the standard protocol, topical pharyngeal anesthesia with 2% lidocaine throat spray was used to alleviate discomfort during TEE. RT3D transesophageal echocardiographic images were obtained in all subjects in the left lateral decubitus position by an experienced sonographer using an x7-2t transducer on a Philips iE33 ultrasound machine (Philips Medical Systems, Andover, MA), which has multiplane 2D, biplane 2D, Doppler imaging, and RT3D capabilities. Synchronized electrocardiograms were recorded and displayed on the images. The transesophageal echocardiographic transducer was placed at the midesophageal level to obtain an optimal biplane view of the ASD. The transducer was adjusted to place the target structure as close to the center of the scan as possible, and RT3D images of ASDs were obtained to ensure that the region of interest would be captured by the most centered scan lines. The 3D zoom mode was used to image en face views of ASDs. The zoom mode displayed a truncated pyramidal data set of variable sizes in both the xy and z directions. The larger the sector width, the lower the temporal resolution, which may drop below 10 Hz when the imaging sector is chosen in both the xy and z directions. To clearly discern the orientation of ASDs and their neighboring structures, the transducer was placed to obtain images of standard transverse sections and the longitudinal plane of TEE as much as possible ( Figure 4 ). The superior vena cava and aorta were also the target structures of zoom mode to be acquired as much as possible ( Figure 4 ). Using the rotation key of the ultrasound machine, RT3D transesophageal echocardiographic images from the left and right atrial perspectives of the interatrial septum for ASDs were obtained online ( Figure 5 ). All echocardiographic data were stored digitally for subsequent offline or online analysis.
All measurements were made in real time at the point of care and the measurements of RT3D TEE were made using online QLAB software (Philips Medical Systems). The diameters of ASDs were obtained in end-systole or early diastole freeze frames, when the ASDs were known to be at their maximal dimensions. All measurements were made by the physician sonographer who operated the system. The major diameters of ASDs with conventional 2D TTE were determined by the maximal dimension of the ASDs in the apical 4-chamber or parasternal short-axis view. The major and minor diameters of en face views of ASDs, on both 2D TEE and RT3D TEE, were determined by the long-axis and short-axis dimensions of the ASDs.
All statistical analyses were carried out using SPSS version 13 (SPSS, Inc, Chicago, IL). Image quality was expressed as the percentage of patients whose images were evaluated consistently by two independent physician sonographers. Nominal variables were compared using χ 2 tests. All continuous variables are reported as mean ± SD. Statistical differences between two groups were calculated using paired Student’s t tests. Correlation coefficients for comparing measurements obtained on 2D TTE and RT3D TEE were calculated. The level of statistical significance was set at P < .05.
En Face Views of ASDs by 2D TTE
Eighty-nine percent of patients (368 of 415) had consistent grading scales of image quality of en face views of ASDs by two independent experienced physician sonographers. The numbers of patients with excellent, good, moderate, and poor image quality were 159 (43%), 135 (37%), 37 (10%), and 37 (10%), respectively. According to the above definition, images of excellent and good quality were considered acceptable. En face views of ASDs on 2D TTE were obtained in 294 patients (80%) in whom secundum ASDs could be adequately imaged in more standard 2D imaging planes. Of the 368 patients, the numbers of those with maximal dimensions of ASDs <2.0, 2.0 to 3.0, and >3.0 cm on conventional 2D TTE in parasternal and apical views were 140, 136, and 92, respectively. Acceptable en face views of ASDs on 2D TTE were visualized in 131 (94%), 112 (82%), and 51 (55%) patients, respectively. The differences were statistically significant (χ 2 = 6.18, P = .045).
Of the 294 patients, 5 had ASD diameters recorded without specific views, and 10 did not show ASDs in the apical 4-chamber view because of the locations of their ASDs toward the superior vena cava. Comparisons of the maximal ASD diameters between the parasternal short-axis, apical 4-chamber, and conventional 2D transthoracic echocardiographic precordial views and the 2D transthoracic echocardiographic en face view were performed in 279 patients with excellent and good image quality ( Table 1 ). There were good correlations among these measurements ( P < .001). The maximal ASD diameter of the conventional 2D transthoracic echocardiographic precordial views came from the parasternal short-axis view in 147 patients (53%).