Dr. Matthew Bramlet describes how 3D printing is being used to improve pediatric cardiac surgery.
Dr. Matthew Bramlet presented a study at the American Heart Association Scientific Sessions in November involving a 9-month-old girl, a 3-year-old boy and a woman in her 20s, all of whom are his patients in Peoria and all of whom were born with complex congenital heart defects.
In each case, surgeons were aided in the planning of reconstructive heart surgery by the creation of 3D heart models from 2D images from the patients, allowing them a view of structural abnormalities that could not be detected in traditional imaging.
Pediatric cardiology is very different than adult cardiology. In its most complex form, we have children who are born with only half a heart. It can take a series of surgeries to get their heart to a point where it can sustain life. Some of those surgeries can be quite complex where we try to re-route the blood flow through these convoluted pathways, which requires the surgeon to know how to put a patch in. As a congenital heart imager, it's my job to provide them with the best information so they are not surprised when they get inside the heart.
I have been able to team up with bioengineers at Jump Trading Simulation & Education Center (www.jumpsimulation.org) to create exact replica models of children’s hearts. It’s been fun to be a part of this project in which we produce a 3D heart model from 2D images. We’ve been on this journey where we are creating processes that have never existed before, and the innovative part of the Jump lab is this mindset that if it doesn’t exist, we will create it.
Creating the Model
Over the past year we’ve been refining our process. We’ve had challenges, but also successes. We’ve now refined it down to where we can fairly quickly take a heart out of an MRI and literally put a model of it into our hands. However, if we're going to make medical decisions on these printed models, which has never really been done before, we have to make sure that we are, in fact, improving our knowledge of these hearts and not producing a false understanding of the anatomy.
We have quality assurance measures and rules that we're applying to the assessment and creation of these models. When we segment the heart, we do it in a very specific manner that replicates the actual anatomy. We go through slice by slice and make sure the model is going to represent the true anatomy. After we perform that and evaluate the heart and make the medical decisions, we always want to take it back to the source image the model came from, and verify our understanding.
Every heart that we've done we've cut out in the same manner. There are two priming chambers, the atria, and the two ventricles. Between those are the atrioventricular valves and two outflows of the ventricles, the semilunar valves. What we've done is cut each heart so that all four valves are retained in that middle critical slice. That allows for several things. Number one, it allows the surgeon to look down through the atrioventricular valve and into the heart in a way that is similar to their actual view during the procedure. Number two, it allows us to look “under the hood”, so to speak, to see what the anatomy is truly like. When you cut the heart in this manner, it's similar to an orientation called a short axis stack, which we use in MRI or parasternal short axis and is a similar format that we use in echo.
By cutting it in this manner on every heart, it's a familiar orientation for both the cardiologist and the surgeons. It allows them to handle the heart with one hand. The easier it is for someone to get through the heart, the less of a barrier it will be to his or her understanding.
Initially, we didn't know the heart models would actually be to scale when holding it in our hands. To test this, I brought in a whiffle ball from my back yard and ran the ball through the CT scanner then rendered it in the same computer software and printed it. You can see, as far as gross scale is concerned, they're identical. This really helped us to verify that when we're holding these hearts in our hand we are holding it at the actual size.
There are some surgeries that are rather straightforward, but there are others that are not. In these complex forms of congenital heart disease, the current methods for evaluating these hearts have included chest x-ray and echocardiography.
Over the past 10-15 years, MRI and CT have begun to play a much larger role in the assessment of complex lesions. Surgeons typically rely on 2D images taken by ultrasound, MRI and CT scans to plan their surgeries. But these images may not reveal the complex structural defects present in a patient.
Now, with 3D printing, surgeons can make even better decisions before they go into the operating room. Using the 2D images to create the 3D models, even out of simple plaster-like materials, has provided us with very valuable information that wasn’t available before. The more prepared they are, the better decisions they make, and the fewer surprises they encounter.
The very first patient case in which we used this specific tactic was really just a proof of concept on a patient with a diagnosis similar to one that was coming up. However, the future case was much more complex. After printing and assessing the heart, I caught the surgeon just before he entered the OR and had him look through the heart. When he was looking through the heart, he said "Oh yeah, this is fantastic, I can see exactly where the patch needs to go. But what about this Swiss cheese VSD?"
An atrial septal defect is a defect in the top two chambers of the heart, and the additional hole is known to avoid detection by traditional methods because with the larger hole there's no pressure gradient driving the force across it. But it was the surgeon’s understanding of the 3D architecture of the Swiss cheese VSD that allowed him to recognize it in the 3D form.
This changed the procedure for the patient. The first thing he did when he went in was confirm that the 3D model was accurate and the Swiss cheese VSD existed. He then altered the surgery. In that case, it saved ischemic time. The heart has to be down to get inside, and it was down less because he confirmed the situation immediately then instituted the back-up plan. When you’re trying to figure out the optimal operation there’s no better way to do so than to actually see the heart and hold it in your hand before attempting the surgery. In the future, this is going to allow us to avoid stopping the heart altogether in that sort of scenario.
The second case in which we utilized this method was a patient who was born with half of a heart but had the raw materials needed for a complete repair. The connections were so twisted that the surgeon had been recommended to go down a single ventricle pathway. He and his team came to us for a second opinion. We reviewed all the images and all the data that they brought with them and initially agreed with that determination. We told them based on these findings this was not something where a two-ventricle repair could work out.
However, we wanted to do an MRI and print the heart to see if it provided a new level of understanding, perhaps allowing us a two-ventricle repair. Sure enough, once I was holding the model my understanding was greatly increased.
In each of these cases, printing a 3D model made a difference in the patient's life. We changed the plan based on the information we gleaned from the model. Since then, we've printed multiple hearts for surgical planning, and with each one we learn something new about the patient. It doesn't always work out to be this dramatic of a change in surgery, but there's a level of understanding that is gained that is simply unparalleled by any other method.
For example, we had a patient that needed a cardiac MRI for other indications, and an ASD happened to be highlighted. In training and practice, we see this hundreds of times in different echoes and modalities. My level of understanding, even for something that I thought I knew inside out, instantly changed when I looked at that model. That is what we want to pass on.
As physicians, we are teachers and educators at heart. Not only are we educating ourselves and our surgeons, we're educating young physicians who are coming up in training. The level of understanding that can be gained from each one of these hearts is tremendous. It's going to change medical education.
We’re also educating families. I'm usually one of the first people to talk to these families. I have to be the one to tell them that their baby has congenital heart disease. It can be very difficult, and families handle it differently. I can go through the anatomy, and go through what is going to be involved in treating their child, and try to explain the situation as fully as possible, but I'm limited in my ability to really show them and have them understand what that anatomy physically looks like.
When I'm talking to a family, I don't just present the problem, I present the solution. With the addition of this congenital heart library, instead of drawing what their baby's heart looks like on a piece of paper, I'm able to actually produce a 3D model heart that is exactly like their baby's, and say, "This is your baby’s heart condition." I’m then able to identify the structures and say, "now see this is the problem – and this is how we are going to fix it."
The better they can understand the problem, the more real – and more helpful – the solution becomes to them. My goal is to have them walk away from our meeting with hope, understanding that everything can and will be ok. The ability to show them with the 3D printed heart is phenomenal.
3D Print Exchange & the National Institutes of Health
While we are using these models for actual surgical planning – which I think is a tremendous, incredible use for our patients – there are also a lot of opportunities for sharing this information with the world in a library format.
An issue we’ve identified is that there aren’t anatomic models that demonstrate congenital heart disease or all its variations. We seek to maximize the 3D modeling resources we have at Jump to offer printable images through the NIH 3D Print Exchange, so that anyone who is willing to learn can do so for free.
We hold ourselves to a very high quality assurance standard because we want people to learn and to use these models as training tools. Our next step is to elevate the project by working with a collaborative group of people and create a peer-review system that provides accurate anatomic examples of congenital heart disease. We are extremely excited about this endeavor and look forward to working with those who are interested in partnering with us.
About the Author
Matthew Bramlet, MD, is Assistant Professor of Pediatric Cardiology and Director of the Congenital Heart Disease MRI Program at the University of Illinois College of Medicine at Peoria.