Ann Thorac Surg 2008;86:273-277. doi:10.1016/j.athoracsur.2007.06.001
© 2008 The Society of Thoracic Surgeons
New Technology
Physical Models Aiding in Complex Congenital Heart Surgery
Sibylle Mottl-Link, MDa,*,
Michael Hübler, MDb,
Titus Kühne, MDc,
Urte Rietdorfd,
Julia J. Krueger, MDc,
Bernhard Schnackenburg, PhDe,
Raffaele De Simone, MDa,
Felix Berger, MDc,
Amy Juraszek, MDf,
Hans-Peter Meinzer, PhDd,
Matthias Karck, MDa,
Roland Hetzer, MD, PhDb,
Ivo Wolf, PhDd
a Department of Cardiac Surgery, University Heidelberg, Heidelberg, Germany
b Department of Cardiac Surgery, German Heart Institute Berlin, Berlin, Germany
c Department of Congenital Heart Disease and Pediatric Cardiology, German Heart Institute Berlin, Berlin, Germany
d Department of Medical and Biological Informatics, German Cancer Research Center, Heidelberg, Germany
e Philips Medical Systems, Hamburg, Germany
f Cardiac Registry, Department of Pathology, Children's Hospital Boston, Harvard Medical School, Boston, Massachusetts
Accepted for publication June 1, 2007.
* Address correspondence to Dr Mottl-Link, Cardiac Surgery, University Heidelberg, Im Neuenheimer Feld 110, Heidelberg, 69120, Germany (Email: sibylle_link{at}med.uni-heidelberg.de).
 |
Abstract
|
|---|
Purpose: Our aim was to improve spatial imagination of complex congenital cardiac abnormalities for subsequent surgical intervention.
Description: Magnetic resonance imaging data of a patient with complex congenital heart malformations was post-processed with software developed at our institution. The resulting virtual surface data sets were printed out three-dimensionally by rapid prototyping techniques.
Evaluation: We present the first patient operated on with intraoperative use of physical models representing the intracardiac volumes (RepliCast) or muscle and vessel walls (RepliCardio). The courses of the coronary arteries were visible on the RepliCast, whereas the RepliCardio showed intracardiac views a surgeon could never obtain intraoperatively in the relaxed heart. Other than on virtual reconstructions presented on computer screens, physical models vastly improve the spatial imagination and give precise information regarding localization and actual size of abnormal structures. The self-explanatory utility of these models shortened preparation and expedited orientation on the open heart.
Conclusions: The additional spatial information provided by RepliCast and RepliCardio models may enable even high-risk correction procedures in patients with complex congenital heart disease.
|
Introduction
|
|---|
Congenital heart surgery remains despite modern imaging techniques such as ultrasound, magnetic resonance imaging (MRI), or computed tomography as one of the most challenging fields in modern medicine. This is chiefly due to the great variety of unique complex cardiovascular morphology. On intraoperative inspection, the open and cardioplegic heart lays distorted in the thorax, very much unlike physiologic conditions making orientation difficult. In addition, most patients have experienced various surgical interventions leading to extensive scarring.
Until now, intraoperative three-dimensional (3-D) spatial imagination of complex congenital heart disease has been based on two-dimensional imaging data and is highly dependent on the experience and skills of the surgeon. Virtual 3-D reconstructions on computer monitors may be helpful when examined with red-green or red-blue glasses, shutter glasses, or special 3-D displays (eg, see www.dresden3d.com; accessed May 19, 2008) [1, 2]. However, orientation is easily lost in virtual space as the human brain recognizes 3-D shapes of familiar objects but may fail to do so if objects do not match accustomed shapes [3]. In addition, the actual comprehension of object sizes is lost due to zoom effects.
Spatial sense can be improved by the tactile qualities of solid models [4]. Models can be produced by several different rapid prototyping techniques including stereo lithography, laser sintering, and 3-D printing [5, 6]. Until now, the main applications of rapid prototyping in clinical settings have involved the human skeleton with benefits for diagnosis and a treatment plan, as well as a reference during the operation [7, 8].
Unlike bone, the living heart does not delineate sharply from surrounding tissue and it is constantly moving. Therefore, the first examples of 3-D presentations developed at our institution showed cardiac morphology of pathologic postmortem specimens [9, 10]. However, a living heart could not be replicated in adequate quality [11] because unlike pathologic specimens, the living heart muscle is not surrounded by air and does not yield a good contrast to neighboring structures. Improved software on automatic endocardial and endoluminal border detection in MRI or computed tomography data sets [12] enabled the presentation of the first cast (RepliCast) of a living patient [13]. This ability to produce a cast was the initial requirement for an even more challenging goal (ie, the production of models of living hearts showing muscle as solid material surrounded by air [RepliCardio], similar to the appearance of a pathologic specimen). However, comparable with pathology, the exact control of curved cut lines giving a view into the closed cavernous virtual surface models was mandatory for the production of RepliCardios.
|
Technology
|
|---|
RepliCast (Fig 1) and RepliCardio (Fig 2) models were produced based on 3-D MRI data (Fig 3). The first patient operated on with additional intraoperative orientation using physical models was 24 years old and had the diagnosis of transposition of the great arteries, pulmonary atresia, large ventricular septal defect, atrial septal defect, tricuspid regurgitation, and dextrocardia. He had had several surgical interventions including aortopulmonary anastomosis after birth, at the age of 6 and 7 years, and dilatation of the pulmonary bifurcation at 10 years of age. The ethics committee approved the study and the patient gave informed consent.
View larger version (112K):
[in this window]
[in a new window]
|
Fig 1. RepliCast model from the (A) right posterior, (B) left anterior, (C) right anterior, (D) cranial, and (E) basal views. There is atrioventricular discordance with connections of the right atrium (RA) to the morphologically left ventricle (LV) on the right side. The left atrium (LA) with an enlarged left atrial appendage (LAA) is connected with the morphologically right ventricle (RV) on the left side. A dilated aorta (Ao) rises from the right ventricle. The left coronary artery (LCA) and right coronary artery (RCA) rise from a common origin of the coronary arteries (CO). (A = anterior; L = left; LAD = left anterior decending coronary artery; LCX = left circumflex; P = posterior; R = right.)
|
|
View larger version (104K):
[in this window]
[in a new window]
|
Fig 2. (A) The RepliCardio model consists of two pieces (B and C). A single cut divides the right atrium (RA) and the left ventricle (LV), allowing the view on the mitral valve annulus (MVA), two papillary muscles (PM), the left anterior descending coronary artery (LAD), ventricular septal defect (VSD), ventricular septum (VS), atrial septal defect (ASD), and the aorta (Ao). (A = anterior; L = left; P = posterior; R = right.)
|
|
View larger version (68K):
[in this window]
[in a new window]
|
Fig 3. Reconstructed three-dimensional magnetic resonance imaging data set of the patient showing the views in three dimensions (lower right: positions of the transversal, vertical and sagittal slice within the three-dimensional data set; upper left: transversal image slice; lower left: vertical image slice; upper right: sagittal image slice).
|
|
|
Technique
|
|---|
A 3-D whole-heart, navigator-corrected MRI data set with nearly isotropic voxel size was acquired using a 1.5 Tesla scanner (Philips Intera CV, Best, NL; Intera 1.5T, Philips Medical Systems, Hamburg, Germany) during free breathing. The data set consisted of 175 slices with a resolution of 1.2 mm (reconstructed, 0.7 x 0.7 x 0.9 mm). The following sequence measurements were used: electrocardiographic gating, slice orientation = axial, single phase, field of view = 360, matrix = 304, repetition time = 4.6 ms, echo time = 2.3 ms, flip angle = 100°, sensitivity encoding (SENSE) factor = 2.2, T2-preparation pulse (echo time = 50 ms)(Brittain), and fat saturation. The acquisition took approximately a maximum of 8 minutes.
Software tools that enable initial segmentation in MRI, computed tomography or live-3-D echocardiography data sets, 3-D rendering, and virtual positioning of cut lines were developed at our institution.
The epicardium was segmented manually with a software tool that allowed automatic interpolation between the slices. The initial segmentation process of the epicardium and great vessels took 9 minutes. Endocardial and endoluminal border detection was performed by choosing a grey scale threshold. The virtual RepliCast surface data were available after approximately 15 minutes from the start of segmentation. The final virtual results (Fig 4) could be controlled with red-blue glasses before 3-D printing.
View larger version (48K):
[in this window]
[in a new window]
|
Fig 4. Virtual surface of the (A) RepliCast and (B) RepliCardio, which can be viewed with red-blue glasses for optimal virtual imagination.
|
|
For the RepliCardio model, the data of the virtual cast was subtracted from the epicardial segmentation that resulted in a closed cavernous surface model. Accurate positioning of curved cut lines or fenestrations into the virtual surface model was mandatory for the production of RepliCardios. As these were initial experiences as to where to place the cut lines, it took several hours and seven trials before a satisfactory virtual RepliCardio surface model was created. However, this amount of time is expected to be reduced significantly with increasing experience.
The 3-D-printer ZPrinter 310 (Z Corporation, Burlington, MA) created numerous thin layers of powder based on plaster and using an original ink jet print head to jet a binder down on the powder. At the end of the building process the remaining loose powder is vacuumed off leaving the final solid model. The RepliCast was manufactured by the laser sintering technique. The cost for the laser sintering model was approximately $810 (600 Euro) and the 3-D printing was approximately $364 (270 Euro). The printing process took approximately 7 hours. For future models, and presuming optimal logistic conditions, assuming MRI acquisition in the morning, the models could be ready the afternoon of the following day.
|
Clinical Experience
|
|---|
There was a clear advantage of the physical printouts represented by the RepliCast (Fig 1) and RepliCardio (Fig 2) for the preoperative virtual 3-D presentations (Fig 4). Spatial imagination of the exact positions and sizes of highly complex abnormal structures were simplified by direct comparison between physical models and the living heart. Other than the beating or cardioplegic heart, the RepliCast and RepliCardio models were able to present formerly unobtainable insights and they provided information about blood-filled cavities hidden behind myocardial walls. The sizes of the models were identical to those of the heart of the patient. Unlike former practices, when surgeons operated with the remembrance of preoperative images, the physical models were taken into the operating room where they were used for direct orientation at the opened situs. The RepliCast (Fig 1) was extremely useful for the initial preparation procedure in representing blood-filled cavities, because it showed the course of the coronary arteries even more precisely than "real" casts made in anatomy or pathology [14]. Surgical preparation is considered high risk for patients who were previously operated on, because vessels can be torn and cavities can burst while scars are dissected. In this case, the dissection of scarred adhesions was guided by immediate direct comparison of the open situs with the RepliCast. Thus, pathologic localization of the coronary arteries (as well as geometry of highly abnormal structures) was a lot easier.
The large ventricular septal defect was intraoperatively closed. The left-sided morphological right ventricle was connected with the pulmonary artery. A valve-carrying conduit was implemented between the myocardial wall of the right-sided morphological left ventricle and the ascending aorta. This procedure would not have been risked without the additional intraoperative orientation on the RepliCast model, which showed the exact localization of the left coronary arteries. If this information were only presented preoperatively on two-dimensional or even 3-D imaging data it could not be recalled intraoperatively, even by highly experienced surgeons. Only direct comparison between the model and the intraoperative situs could help in this high-risk procedure. Postoperatively the hemodynamic condition was stabilized. However, unfortunately the patient died 7 months after the operation due to sepsis.
|
Comment
|
|---|
In comparison to 3-D presentations, which are becoming routinely available for MRI, CT scans, and even echocardiography, our 3-D visualizations (Fig 4) can be additionally viewed with red-blue glasses or special 3-D computer screens for optimal spatial imagination. However, human sense can easily be deceived, especially concerning unfamiliar objects perceiving, for example, a human nose in front of the ears, even when a model shows it vice versa. Therefore, even the most experienced surgeons have difficulties in exactly localizing abnormal structures in complex congenital heart disease in which every case is unique. In addition, the memory of spatial information between preoperative planning and intraoperative proceeding is highly difficult in these cases, if not impossible. Therefore, the RepliCast (Fig 1) was much more useful for intraoperative preparation than its virtual match (Fig 4).
The RepliCardio model expedited orientation on the open heart. The exact position of the ventricular septal defect could be localized more posterior and inferiorly than had been expected. The advantage of the RepliCardio was that it reflected views that could never be intraoperatively obtained. Other than a cardioplegically distorted heart, the RepliCardio showed physiologic conditions. In addition, virtual cuts could be placed where no surgeon would dare to open the cardiac muscle.
However, a limitation was that the heart valves and the thin atrial septal wall were too delicate for image acquisition, despite the high resolution of the magnetic resonance images. Ultra-fast dual camera 64-detector computed tomographic scans might be more appropriate than MRI for the final production of structures with extreme thinness and high mobility. Our software is also suitable for use in computed tomographic data sets. In addition, interactive local threshold adaptation similar to the technique presented on pathologic specimens could be used to compound delicate textures, such as the heart valves and cordae tendinea in computed tomographic data sets [10].
Another limitation was that the models were not sterile; therefore, the surgeon was not able to touch them intraoperatively. In our case, the surgeon was dependent on his assistant to show the models to him at angles he wished to see, which was unnecessarily time consuming. If models were made out of metal (eg, titanium), then they could be sterilized similar to other surgical instruments. Another optional material would be elastic rubber, which would allow flexible distension of the fenestrations for improved insight. In addition, even cutting into the models with scalpels would be possible. Thus, surgeons would be able to place their own fenestrations in a manner they are used to in the operating room. Furthermore, 3-D printouts could be manufactured in color, which might be useful for the marking or labeling of important structures.
In summary, the RepliCast model was extremely useful for the initial preparation procedure, whereas the RepliCardio model improved orientation at the open heart. Therefore, physical models replicating the inner volumes or muscle and vessel walls are expected to greatly simplify complex congenital heart surgery.
|
Disclosures and Freedom of Investigation
|
|---|
Software was developed within the special research area 414 (computer and sensor-guided surgery), which was financed by the Deutsche Forschungsgemeinschaft (German research association). Amy Juraszek is supported by grant no. P50 HL074734-01. The authors had full control of the design of the study, methods used, outcome measurements, analysis of data, and production of the written report. The tested technology was not purchased.
|
Acknowledgments
|
|---|
The authors thank Dr Aoife Hunt for proofreading the article.
|
Footnotes
|
|---|
Disclaimer The Society of Thoracic Surgeons, the Southern Thoracic Surgical Association, and The Annals of Thoracic Surgery neither endorse nor discourage use of the new technology described in this article.
|
References
|
|---|
- Ribas GC, Bento RF, Rodrigues Jr AJ. Anaglyphic three-dimensional stereoscopic printing: revival of an old method for anatomical and surgical teaching and reporting J Neurosurg 2001;95:1057-1066.[Medline]
- Sorensen TS, Pedersen EM, Hansen OK, et al. Visualization of morphological details in congenitally malformed hearts: virtual three-dimensional reconstruction from magnetic resonance imaging Cardiol Young 2003;13:451-460.[Medline]
- Bülthoff I, Bülthoff H, Sinha P. Top-down influences on stereoscopic depth-perception Nat Neurosci 1998;1:254-257.[Medline]
- Way TP, Barner KE. Automatic visual to tactile translation—part II: evaluation of the TACTile Image Creation System IEEE Trans Rehabil Eng 1997;5:95-105.[Medline]
- Winder J, Bibb R. Medical rapid prototyping technologies: state of the art and current limitations for application in oral and maxillofacial surgery J Oral Maxillofac Surg 2005;63:1006-1015.[Medline]
- Seitz H, Rieder W, Irsen S, et al. Three-dimensional printing of porous ceramic scaffolds for bone tissue engeneering J Biomed Mater Res B Appl Biomater 2005;74:782-788.[Medline]
- Erickson DM, Chance D, Schmitt S, et al. An opinion survey of reported benefits from the use of stereolithographic models J Oral Maxillofac Surg 1999;57:1040-1043.[Medline]
- D'Urso PS, Askin G, Earwaker JS, et al. Spinal biomodeling Spine 1999;24:1247-1251.[Medline]
- Vahl CF, Meinzer CF, Hagl S. Three-dimensional presentation of cardiac morphology Thorac Cardiovasc Surg 1991;39(Suppl 3):198-204.[Medline]
- Wolf I, Makabe M, Greil G, et al. Image processing methods for an exact reproduction of unique waxed heart specimensIn: Meiler M, Saupe D, Kruggel F, Handels H, Lehmann T, editors. Bildverarbeitung in der Medizin 2002. Heidelberg, Germany: Springer; 2002. pp. 33-36.
- Binder TM, Moertl D, Mundingler G, et al. Stereolithographic biomodeling to create tangible hard copies of cardiac structures from echocardiographic data: in vitro and in vivo validation J Am Coll Cardiol 2000;35:230-237.[Abstract/Free Full Text]
- Wolf I, Vetter M, Wegner I, et al. The medical imaging interaction toolkit Med Image Anal 2005;9:594-604.[Medline]
- Mottl-Link S, Boettger T, Krueger JJ, et al. Cast of complex congenital heart malformation in a living patient Circulation 2005;112:356-357.
- Anderson RH, Becker A. Anatomie des Herzens (Cardiac anatomy)Thieme: Stuttgart; 1982.
This article has been cited by other articles:
|
|
|
|
 
E. Riesenkampff, U. Rietdorf, I. Wolf, B. Schnackenburg, P. Ewert, M. Huebler, V. Alexi-Meskishvili, R. H. Anderson, N. Engel, H.-P. Meinzer, et al.
The practical clinical value of three-dimensional models of complex congenitally malformed hearts.
J. Thorac. Cardiovasc. Surg.,
September 1, 2009;
138(3):
571 - 580.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Kalejs and L. K. v. Segesser
Rapid prototyping of compliant human aortic roots for assessment of valved stents
Interactive CardioVascular and Thoracic Surgery,
February 1, 2009;
8(2):
182 - 186.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
