Cardiac Axis-Oriented Full-Volume Data Acquisition in Real-Time Three-Dimensional Transesophageal Echocardiography to Facilitate On-Cart Analysis : Anesthesia & Analgesia

Secondary Logo

Journal Logo

Cardiovascular Anesthesiology: Technical Communication

Cardiac Axis-Oriented Full-Volume Data Acquisition in Real-Time Three-Dimensional Transesophageal Echocardiography to Facilitate On-Cart Analysis

Hirasaki, Yuji MD*; Seino, Yusuke MD; Tomita, Yuko MD; Nomura, Minoru

Author Information
doi: 10.1213/ANE.0b013e3182294b91

Real-time 3-dimensional transesophageal echocardiography (3DTEE, iE33, Philips Medical Systems, Andover, MA) is increasingly used in the setting of cardiac surgery. A growing body of evidence suggests the advantage of real-time 3DTEE over conventional 2DTEE for the diagnosis of cardiac pathologies and for decision making in a variety of clinical settings.16 The usefulness of 3DTEE is particularly noteworthy in mitral valve (MV) repair. Mitral regurgitation derives from a complex spectrum of dysfunctions. From a research standpoint, 3DTEE reveals aspects of mechanics and physiology useful for developing new approaches to repair.7,8 From a clinical operations perspective, complex leaflet defects are more readily appreciated by experts and nonexperts alike when observed in 3D.

In the current real-time 3DTEE, both conventional 2D and 3D images can be obtained using the same matrix array probe (X7-2t, Philips). Because these 2 modes differ with respect to handling of the datasets, new basic knowledge and skills are required for echocardiographers to efficiently manipulate the 3D image dataset intraoperatively. Full-volume (FV) 3D imaging is a mode in which 4 to 7 wedge-shaped, gated subvolume datasets are obtained and reconstructed into a large pyramid-shaped 3D image displayed at a relatively high frame rate.4 A single FV dataset contains more information than that obtained from other real-time 3D imaging modes. The FV dataset can be used to analyze left ventricular (LV) volume and MV geometry using plug-in software (QLAB and MVQ, respectively; Philips) installed on the echo machine.6 Information of multiple aspects can be obtained by freely rotating the dataset to view from any orientation. Additionally, valuable information can be obtained by cropping unwanted portions of the 3D dataset using 3 pairs (green, red, and blue) of planes that move along the 3D axes (fixed box cropping) or a plane that can be placed in any desired position (arbitrary plane cropping).3 An important feature of the current real-time 3DTEE system is that the orthogonal axes in each 3D image are set automatically according to the X-Y axes of the initial multiplane 2DTEE image, and therefore axis orientation of the 3D dataset differs depending on the initial 2D plane angle. See, for example, the difference in orthogonal axes and box-crop plane alignment between FV datasets obtained with the midesophageal (ME) 4-chamber (4C) view (panels A and B) and those obtained with the ME mitral commissural (MC) view (panels C and D) in Figure 1.

F1-11
Figure 1:
A, Perspective image of the uncropped FV 3-dimensional (3-D) dataset acquired with the midesophageal 4-chamber (ME4C) view as the initial 2D plane (4CFV) with orthogonal axes (narrow lines) and box crop planes (bold lines). B, The 4CFV dataset rotated to face the blue-top crop plane showing a simulated image of the heart as viewed from the atria (transatrial heart view). The ME4C is located in the center of the dataset (white line). C, Perspective image of the uncropped full-volume 3D dataset acquired with the MEMC view as the initial 2D plane (bicommissural full-volume, BCFV). D, The BCFV dataset rotated to show the transatrial heart view. The MEMC plane is located in the center of the dataset (white line). AV = aortic valve; ME4C = midesophageal 4-chamber plane; MEMC = midesophageal mitral commissural plane; MV = mitral valve; PV = pulmonic valve; TV = tricuspid valve; D = dimensional; FV = full volume.

Intuitively, most echocardiographers acquire the initial FV3D datasets with the ME4C view, because it is a “gold standard” 2DTEE image.9 Other analogues of standard 2DTEE images, however, are difficult to obtain using this 3D dataset. In the 4CFV dataset, box crop planes are aligned to “obliquely” cross the MV or aortic valve (AV) (Fig. 1B). Except for the ME4C and ME2C views, the sectional images obtained using the box-cropping planes are not clinically relevant (Fig. 2). Although analogues of standard 2DTEE views can be obtained using the arbitrary plane-cropping process, the process is time consuming and the image orientation of the cropped images is difficult for other echocardiographers or trainees to envisage. Because of the complexity involved in handling the image datasets, the information contained within a single dataset is not fully used. Instead, multiple 3D datasets concerning the same cardiac structures within a variety of orthogonal axes are obtained as part of a single 3DTEE study. This is quite time consuming and imposes an additional workload on the perioperative echocardiographer. There has been no standardized scanning method for real-time 3DTEE. We therefore developed a comprehensive acquisition approach, bicommissural FV (BCFV) data acquisition, to maximize the information obtained about the cardiac anatomy within the dataset and to simplify the complex cropping process.

F2-11
Figure 2:
Cropped images of the 4-chamber full-volume dataset obtained using the green (A, B, and C) and the right-red box crop planes (D, E, and F). AV = aortic valve.

METHODS

Our BCFV acquisition can be accomplished as follows:

  1. Obtain a 2D ME MC view.
  2. Press the Full volume button in the console, and a preview of the 2D ME MC and LV long-axis views are displayed in the left and right panels, respectively (Fig. 3).
  3. Adjust the scan area by turning the Image priority knob to include the AV and the entire MV. Scan depth can be set to include the whole LV (complete BCFV) or reduced to the level of the papillary muscles (shallow BCFV).
  4. Adjust the probe position (e.g., retroflex) so that the LV apex is aligned to the center of the sector.
  5. Acquire FV data, which displays as a 3D image oriented in the initial plane. Press the Reset crop button to display an uncropped FV image.
F3-11
Figure 3:
Preview of the bicommissural full-volume data acquisition showing the scan area and alignment of the orthogonal axes (red, blue, and green lines). The X-Y axes are set as those of mitral valve commissural view (left panel), and the subvolumes are aligned as the dashed lines along the Z-axis perpendicular to the bicommissural plane (right panel).

RESULTS

In the uncropped BCFV dataset, the heart is aligned in the crop box with en face view of the AV and MV through the front (green) and top (blue) cropping planes, respectively (Fig. 1, C and D). Furthermore, box cropping of the dataset by turning the plane adjustment knob yields many useful sectional images in addition to the analogues of standard 2DTEE views (Fig. 4, Table 1). The 3 pairs of cropping planes can be used as “windows” through which the 3D cardiac anatomy can be efficiently analyzed from a single 3D dataset. Cropping the dataset using the red planes shows sagittal cuts of mitral anterior–posterior leaflets, which allows for assessment of the coaptation of the anterior and posterior segments of the mitral leaflets (Fig. 4). The shallow BCFV dataset provides images of the MV and AV at a higher frame rate and is also useful for quantifying the 3D variables of the MV structure using MVQ software. The reference points required for MV modeling (anterior, posterior, anterolateral, and posteromedial edges of the mitral annulus, papillary muscles, and the AV) are prealigned in the windows of the multiplanar reformatting display (Fig. 5).

F4-11
Figure 4:
Cropped images of the complete bicommissural full-volume dataset. Images in the left column indicate the location of each box-crop plane (panels A–L). A–C, Images obtained using the green front-crop plane. A, Right ventricular inflow–outflow view. B, Short-axis view of the left ventricular outflow tract. C, Mitral commissural view. D–F, Images obtained using the right-red crop plane. D, Cut image of the right heart demonstrating the right ventricle and coronary sinus. E, Left ventricular long-axis view (LVLAX). F, Saggital section of the mitral A1–P1 segments. G–I, Images obtained using the left-red crop plane showing the “mirror images” of panels D–F. J–L, Images obtained using the blue bottom-crop plane showing the left ventricular short-axis view (LVSAX) at the level of the apical (panel I), mid (panel K), and basal (panel L) segments, respectively. A1–A3 = segments of the anterior mitral leaflet; APM = anterior papillary muscle; AV = aortic valve; CS = coronary sinus; IVS = interventricular septum; LV = left ventricle; MV = mitral valve; P1–P3 = segments of the posterior mitral leaflet; PAC = pulmonary artery catheter; PPM = posterior papillary muscle; RV = right ventricle; TV = tricuspid valve.
T1-11
Table 1:
Images Obtained by Box Plane Cropping from Bicommissural Full-Volume Dataset
F5-11
Figure 5:
Multiplanar reconstruction display for mitral valve quantification using the plug-in software (MVQ). In the bicommissural full-volume dataset, 2-dimensional images containing the reference points (green circles) for mitral valve (MV) modeling (i.e., the MV commissural view, the mitral anterior–posterior view, and the short-axis view of the MV) are prealigned to appear in the upper left, upper right, and lower left panels, respectively.

DISCUSSION

The advantage of this method is that a single dataset can be used with minimal manipulation for simple on-cart analysis of cardiac structures from multiple aspects. Because the direction of each cropping plane's movement is fixed, a comprehensive, reproducible assessment is possible using datasets acquired from different patients. Standardizing the 3D alignment of the heart may allow different echocardiographers to easily understand image orientation in similar 3D datasets. As shown in Figure 1C, the AV is critical to image orientation and is readily apparent in this 3D dataset. The sectional images of the LV system obtained using the green, red, and blue cropping planes by our method can be simulated as the sagittal, coronal, and axial sections of the LV system (thus cardiac axis oriented), respectively. It is interesting to note that the LV system (AV, left ventricle, papillary muscles, and MV) set in our present orthogonal axis alignment appears to be a single apparatus, like a teapot, that has a symmetric property, with the LV long-axis plane as a midline. Our speculation is supported by the fact that images obtained using the 2 red crop planes provide the mirror images of each other (panels E and F versus panels H and I in Fig. 4, respectively). Further anatomic investigation is necessary to validate our speculation.

Despite its usefulness, there are several limitations to the present method. First, the image quality of the FV dataset depends on electrocardiogram gating and thus may not be useful in patients with arrhythmia. Second, this method is not suitable for intraoperative real-time monitoring during surgical procedures because the FV 3D mode is not a “true” real-time 3D image modality. Third, a large part of the interatrial septum is out of the scan area of the BCFV dataset, requiring a different approach for analysis of the interatrial septum. Fourth, our method is exclusively applicable to the current Philips real-time 3DTEE system, which is the only real-time 3DTEE imaging modality currently available, and therefore it is not clear whether our technique can be applied to future real-time 3DTEE imaging modalities. The mitral anterior–posterior and mitral bicommissural lines, however, should be recognized by any cardiac imaging modality as the fundamental axes for evaluating LV morphology.

CONCLUSION

We present a useful method to obtain a single FV dataset that allows for efficient assessment of cardiac structures in real-time 3DTEE. The mitral bicommissural region, mitral anterior–posterior axes, and the line that connects the mitral annular center with the LV apex appear to be the key axes for 3D assessment of the cardiac morphology in a geometry-oriented manner. Further investigation to assess the feasibility of our method across patients with a variety of cardiac diseases is necessary.

DISCLOSURES

Name: Yuji Hirasaki, MD.

Contribution: This author helped design the study and write the manuscript.

Attestation: Yuji Hirasaki approved the final manuscript.

Name: Yusuke Seino, MD.

Contribution: This author helped write the manuscript.

Attestation: Yusuke Seino approved the final manuscript.

Name: Yuko Tomita, MD.

Contribution: This author helped write the manuscript.

Attestation: Yuko Tomita approved the final manuscript.

Name: Minoru Nomura.

Contribution: This author helped write the manuscript.

Attestation: Minoru Nomura approved the final manuscript.

REFERENCES

1. Jungwirth B, Mackensen GB. Real-time 3-dimensional echocardiography in the operating room. Semin Cardiothorac Vasc Anesth 2008;12:248–64
2. Fisher GW, Anyanwu AC, Adams DH. Change in surgical management as a consequence of real-time 3D TEE: assessment of left ventricular function. Semin Cardiothorac Vasc Anesth 2009;13:238–40
3. Salcedo EE, Quaife RA, Seres T, Carroll JD. Framework for systematic characterization of the mitral valve by real-time three-dimensional transeophageal echocardiography. J Am Soc Echocardiogr 2009;22:1087–99
4. Vegas A, Meineri M. Three-dimensional transesophageal echocardiography is a major advance for intraoperative clinical management of patients undergoing cardiac surgery: a core review. Anesth Analg 2010;110:1548–73
5. Gripari P, Tamborini G, Barbier P, Maltagliati AC, Galli CA, Muratori M, Salvi L, Sisillo E, Alamanni F, Pepi M. Real-time three-dimensional transesophageal echocardiography: a new intraoperative feasible and useful technology in cardiac surgery. Int J Cardiocasc Imag 2010;26:651–60
6. Mor-avi V, Sugeng L, Lang RM. Real-time 3-dimensional echocardiography: an integral component of the routine echocardiographic examination in adult patients? Circulation 2009;119:314–29
7. La Canna G, Arendar I, Maisano F, Monaco F, Collu E, Benussi S, De Bonis M, Castiglioni A, Alfieri O. Real-time three-dimensional transeophageal echocardiography for assessment of mitral valve functional anatomy in patients with prolapsed-related regurgitation. Am J Cardiol 2011;107:1365–74
8. Chandra S, Salgo IS, Sugeng L, Weinert L, Tsang W, Takeuchi M, Spencer KT, O'Connor A, Cardinale M, Settlemier S, Mor-Avi V, Lang RM. Characterization of degenerative mitral valve disease using morphologic analysis of real-time three-dimensional echocardiographic images: objective insight into complexity and planning of mitral valve repair. Circ Cardiovasc Imag 2011;4:24–32
9. Shanewise JS, Chung AT, Aronson S, Stewart WJ, Weiss RL, Mark JB, Savage RM, Sears-Rogan P, Mathew JP, Quiñones MA, Cahalan MK, Savino JS. ASE/SCA guidelines for performing a comprehensive intraoperative multiplane transesophageal echocardiography examination: recommendations of the American Society of Echocardiography Council for Intraoperative Echocardiography and the Society of Cardiovascular Anesthesiologists task force for certification in perioperative transesopahgeal echocardiography. Anesth Analg 1999;89:870–84
© 2011 International Anesthesia Research Society