Sign In

Communications of the ACM

ACM News

A Near-Perfect Fit


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
Multi-material physical models of patients aortic heart valves, each with its own unique size, shape, and amount of calcification.

An innovative three-dimensional (3D) printing technique enables heart surgeons to print out a near-perfect replica of a patient's aortic valve for use as a replacement.

Credit: Wyss Institute/MHarvard University

Researchers have developed an innovative three-dimensional (3D) printing technique that enables heart surgeons to print out a near-perfect replica of a patient's aortic valve, in order to use is as a replacement.

Armed with the realistic 3D-printed replica, surgeons have a much easier time finding a good match for the patient's valve among the spectrum of artificial valves offered by a variety of medical manufacturers.

Essentially, surgeons can look at a physical model of their patient's heart valve and compare it to artificial valves offered by various manufacturers, to see which artificial valve offers the best fit.

"Our integrative 3D printing and valve-sizing system provides a customized report of every patient's unique aortic valve shape, removing a lot of the guesswork and helping each patient receive a more accurately sized valve," says James Weaver, a senior research scientist at the Wyss Institute for Biologically Inspired Engineering and a corresponding author on the research.

Adds Ahmed Hosny, lead author of the study, who is currently is an applied machine learning researcher in biomedicine at Dana-Farber Cancer Institute, the software "is Web-based and is serverless. All the computation is done in the client's browser and hence mitigates patient data privacy concerns."

Researchers from Harvard University, the University of Washington, the Max Planck Institute of Colloids and Interfaces, and the VA Puget Sound Health Care System came together to create the innovative 3D printing technique.

"At the core of the personalized medicine challenge is the realization that one medical treatment will not serve all patients equally well, and that therapies should be tailored to the individual," says Donald Ingber, founding director of the Wyss Institute.  "It is exciting to see how our community is innovating in this space and attempting to translate new personalized approaches from the lab and into the clinic."

Researchers say the new tool, which needs additional development before it can be commercialized, represents a major step in reducing the risk associated with replacing a human heart valve with an artificial one. Heart valve surgeries, for example, sometimes result in the implantation of valves that turn out to be too small for patients, which can lead to leaking, or even the implanted valve becoming dislodged from the heart.

Moreover, when artificial heart valves turn out to be too large, they can rip through a patient's heart, sometimes resulting in the patient's death.

Fortunately, the new 3D printing technique, once fully developed, could change all that.

"This could become state of the art," says Rob MacLeod, a professor of biological engineering and cardiovascular medicine at the University of Utah.  "I'm very excited about the future of this system."

The 3D printing technique works by relying on traditional methods of three-dimensionally imaging a human organ, then using a software tool created by the researchers to enhance a tough-to-detail facet of heart valves.

Specifically, the researchers begin with conventional X-rays of a patient's heart valve, which are run through imaging programs 3D Slicer and Autodesk Meshmixer to generate a conventional 3D image of the valve. Then the researchers use special software they created to enhance the level of detail on the patient's heart valve 'leaflets'.

That extra step of getting the leaflets imaged just right is critical to generating an accurate 3D image of the heart valve, Hosny says.

As a result, researchers are able to generate an enhanced 3D image of the patient's heart valve—complete with an enhanced level of detail on the heart valve leaflets—which can be printed out on any conventional 3D printer formatted for .stl files.

Further aiding in the matchmaking process is a 3D-printed 'sizer' device the researchers developed.  This fits inside the 3D-printed model of the patient's heart valve, and can be expanded and contracted to better determine the optimum artificial replacement valve for the patient.

The researchers admit that their tool needs more development before it can be commercialized.  In a study involving 30 heart surgeries, for example, their 3D imaging system was able to accurately predict whether a heart valve would be a good fit just 66% of the time.

Agrees Udayabhanu M. Jammalamadaka, a staff scientist at Washington University in St. Louis specializing in 3D medical printing, "Healthcare firms need to make significant modifications before commercializing this product. , For clinical adoption of this protocol, the sensitivity and precision needs to be improved."

Even so, the researchers believe their technique is an excellent first start, and they have generously posted their 3D enhancement software on the Web ( at http://ahmedhosny.github.io/av-generator/) for other researchers to study and improve upon, along with the open-source code for that software

"We wanted to create open source software that could evolve and adapt to the dynamic challenges faced by interventional cardiologists and surgeons," says Beth Ripley, a member of the research team and a radiologist and senior innovation fellow at the University of Washington   "We look forward to seeing iterative changes in the software and protocol in the hands of end-users, as well as a broadening of its applicability to patient-specific procedures."

Adds Hosny,  "We hope the research community can make use of it and improve it, either by contributing code on Github or reporting bugs. Ultimately, we encourage other researchers to take their software pipelines and make them accessible to other non-technical researchers through a simple Web interface or the like. This will have a great impact on accessibility by the larger research community, as well as the reproduciblity of works moving forward." 

Milan Sonka, co-director at the Iowa Institute for Biomedical Imaging at University of Iowa (https://medicine.uiowa.edu/eye/profile/milan-sonka), describes the research as "intriguing."  Ultimately, improving on the tool may involve migrating the process to virtual reality (VR).

The reason:  In addition to rendering in 3D, VR is also able to simulate the beating of a human heart – something that 3D-printed models of organs currently cannot replicate, Sonka says.

"It is only a question of time and additional development of user-friendly virtual reality approaches for that approach to surpass that of 3D printing," Sonka says.

Joe Dysart is an Internet speaker and business consultant based in Manhattan, NY, USA. 


 

No entries found