University of Colorado Hospital Puts Patient Hearts into Physicians Hands

	Routine MR images of the heart of an infant with complex congenital heart disease were transformed by the 3D Lab at the University of Colorado into an accurate physical model. Images courtesy of Joseph Kay, MD, University of Colorado Hospital, Aurora, Colo.

A process called “rapid prototyping” allows interventionalists to produce 3D models of patients’ hearts to better facilitate pre-procedural planning for catheter-based structural heart interventions.

Model plan

Previously limited to engineering and manufacturing, rapid prototyping initially helped create physical models of cars, planes and computers before their designs were sent to the production floor. The technique first entered the medical field to help surgeons—oral and maxillofacial, orthopedic and neurologic—with hands-on approaches to complex procedures. Three years ago, researchers from the University of Colorado Denver School of Medicine and physicians from the University of Colorado Hospital (UCH) in Aurora, Colo., began refining the technique for the cardiovascular realm.

“Rapid prototyping offers us an opportunity to take a patient’s medical image and transform it into a physical object,” says lead researcher John D. Carroll, MD, an interventional cardiologist and medical director of the Cardiac and Vascular Center at UCH.

Carroll works alongside other cardiologists, computer scientists and research fellows in the clinical science program, a collaboration with Philips Healthcare that combines clinical research and developmental approaches to 3D visualization.

While 2D and 3D computer images obtained from coronary CT scans have made great strides in visualizing heart conditions, these images can sometimes “misrepresent many of the features of the disease state of a patient,” says Carroll. Three-dimensional models help to better understand individual features and to provide better planning methods for the navigation of devices during an intervention (see sidebar).

UCH has created 40 patient heart models and is refining the process by which the models are handcrafted. Currently, the facility focuses on working with the most complex patient cases and those that deal with structural heart disease. These cases include ventricular septal defects with surgical patch dehiscence, fenestrated artrial septal defects with large atrial septal aneurysm and prosthetic mitral valve perivalvular leaks.

“Creating these models makes the most sense for patients who have unusual or complex anatomy that needs to be more fully understood before a therapeutic intervention,” says Carroll.

Research Assistant Adam R. Hansgen, adds, “When you look at a model instead of an image on a screen, you start understanding the anatomic relationships better and how you would potentially manipulate the catheter.”

The researchers say that models such as these may pave the way for medical device companies to customize catheter-based devices based on individual anatomic structure.

Putting it together

UCH is one of a handful of facilities in the U.S. working to create prototyped models of the whole heart, says S. James Chen, PhD, director of the 3D Lab and associate professor of medicine. “It’s a pretty small group, and most other researchers doing this type of work are looking only at specific chambers or vessels, not the whole heart.”

The intricate process begins with 2D cardiac CT angiography images (MRI and echocardiography images are sometimes used), which are then processed with the use of several software packages ranging from those used for DICOM viewing to others that were originally designed for applications in computer animation and video game creation.

The 3D datasets are broken down into cross-sectional slices of varying thickness, depending on the technology and model of the 3D printer being used. UCH researchers use a Z Corporation 3D printer equipped with Hewlett-Packard print heads. The heart’s cross-sections are printed from the ink-jet print head using a water-based adhesive. Researchers then use light adhesive and powder made of plaster or resins to coat and glue the model together, creating a porcelain-type feel, says Hansgen.

The model goes into its post-processing phase where it will cure and dry. The 3D printing process takes six to seven hours to complete, while the entire rapid prototyping process—imaging to complete model—takes three days. The plaster or resin-based powders allow color to be injected into the rapid prototyped model to create the contrast and contours of the heart’s elaborate anatomy.

The 3D printing and modeling process is costly and time-consuming. A conventional 3D printer costs between $30,000 and $50,000, and resins, powders and elastomeric materials also can fuel costs. In addition, the heart’s complex structure and amount of soft tissue often make it challenging to recreate.

Rapid Prototyping Case Study: Congenital Muscular VSD

A physician model of the heart of a patient with mitral valve (red) disease was created from a standard CT aquisition. The model, created through a process called rapid prototyping, can be made with removable sections allowing inspection of the inner heart chambers. A 30-year-old man was referred to the University of Colorado Hospital in Aurora, Colo., for percutaneous closure of a congenital muscular ventricular septal defects (VSD). Researchers at the University of Colorado Denver School of Medicine, led by John D. Carroll, MD, created a rapid prototype 3D physical model of the patient’s heart that “clearly defined the VSD anatomy and spatial orientation to surrounding structures” (Circulation 2008;117;2388-2394). Cardiac CT angiography data, which was used to create the rapid prototype, showed the VSD to measure 14x12 mm. Using the 3D model, researchers determined that a 12-mm Amplatzer (AGA Medical) muscular VSD occluder device would not interfere with aortic, mitral or tricuspid valve function. Researchers then simulated placement of various catheters in the model and found the optimal approach would be with a Judkins right-4 diagnostic catheter from the retrograde aortic approach. They determined that this approach would result in minimal catheter bending or kinking, while maximally utilizing the catheter’s primary and secondary bends. The patient subsequently underwent an uncomplicated percutaneous VSD closure through the right internal jugular vein approach and a post-VSD closure left ventricular angiogram demonstrated no evidence of left to right flow.

Modeling for better care

The next phase in the process will be incorporating the tool into device design and development, says Carroll. In fact, UCH has begun to collaborate with the FDA’s Division of Cardiovascular Devices to work on a project that could lead to “more standardization of pre-clinical testing” and “update the process of developing, testing and evaluating new cardiovascular devices,” he says.

UCH also has begun working with six or more device manufactures including Medtronic, AGA Medical, Cook Medical and W.L. Gore & Associates to develop cardiac devices as bench-testing models for training purposes, Chen says. An important next step is developing a utility that can measure the size of the device needed to fit inside a certain heart defect. The use of simulators could help cardiologists test or train how to implant a catheter device directly into the target location within the heart.

Researchers also see the technology as a valuable training and teaching tool, as 3D models can eliminate the need to use autopsy hearts, which can often be limiting because of their tendency to fall apart after use, strong odor and unavailability for wide-spread clinical use. In addition, the 3D models and patient cases become part of a heart library that will provide the facility with a tool to further the design and testing of cardiac devices. The library would also add to teaching methods at the facility as it would enable individual patient cases to be printed and analyzed for instruction.