Project Spotlight: 3D Printed Biomaterials

Drawing on knowledge from the fields of Engineering, Cell Biology, and Physiology, bioprinting is a process by which cells, scaffolds, and extracellular materials can be printed in a customizable fashion. Printed constructs can range from simple 2D models to complex structures that recapitulate native environments. The need for bioprinted implants and organs is significant. At any given time, nearly 3,500 – 4,000 people are waiting for a heart or heart-lung transplant, and every 10 minutes a new person is added to the national transplant waiting list. Based on hierarchical machine learning (HML), this project will enable the manufacturing of high-fidelity 3D constructs from various starting materials, including biological components. The tool will be used by designers, manufacturers, and others who require the ability to integrate information from mathematical models of materials and manufacturing processes in order to leverage 3D bioprinting for efficient and accurate manufacturing.

Please share a little about your background and your research experiences.

Newell Washburn has been a professor at Carnegie Mellon University (CMU) for nearly 16 years, where he teaches chemistry and biomedical engineering. Jennifer Bone received a BS in both Physics and Molecular Biology from UC Berkeley as well as a Master’s in Biomedical Engineering from CMU. She is currently pursuing her PhD at CMU in Dr. Washburn’s research group. In addition to their academic work, both Dr. Washburn and Jennifer have experience with startups and entrepreneurship. Dr. Washburn is a co-founder of Ansatz AI and Jennifer held a position at startup Lucira Health, formerly known as Diassess Inc., before beginning her doctoral studies.

What led you to the PHDA?

There was a call for proposals from the Center for Machine Learning and Health (CMLH) at CMU and we thought the PHDA was an interesting mechanism for support. Our lab is already oriented toward translational science – we explore applications from the outset and then carry out research toward those aims. The structure of the PHDA seemed to be a good fit and was a nice opportunity for us to see how broadly applicable our project and approach would be for systems that are much more clinically relevant than anything we had done yet.

Walk us through your project.

3D Bioprinting ImageOur team is interested in designing implants and transplants that could be individualized to each patient. We know that 3D bioprinting is a good strategy for this type of work due to the level of personalization that is possible. For example, different cells or extracellular matrix components can be printed in any number of combinations and orientations. However, when compared to materials like plastics, 3D printing of biomaterials poses certain challenges. Inherent complexities of biological materials can lead to high variability in end products, and low reliability that a given construct will be biologically functional. It would not be practical to perform the number of traditional tissue engineering experiments needed to better understand these variables. Instead, our team started exploring whether machine learning (ML) could provide the answers. Importantly, our group has already developed algorithms that allow us to make predictions based on small or sparse data sets. That is especially critical in the context of our project, where ML is being applied to a relatively small data set compared with more traditional use cases for ML.

What are your project’s next steps?

We know that when cells are put through the printing process, they are subjected to lots of pressure and sheer stresses, and are then put in an environment that isn’t necessarily ideal. Because we have so many different variables, such as printer type and print method, it’s difficult to know how these will impact the cells and how they will function in the resulting 3D construct. Right now, we are working toward trying to predict the fidelity of printing cellular constructs, specifically evaluating how well our ML techniques can predict cellular outcomes and tissue-level properties. This is a very rich range of objectives for the algorithm to try and optimize. In addition, we’re engaging with companies to explore how our algorithms can complement their products and services.

When you look at Pittsburgh as a region, what role do you see the PHDA playing? What do you foresee the future of innovation looking like here?

Every industry is going to be transformed by data science, artificial intelligence (AI), and ML. This won’t only impact existing industries, but also industries of the future like artificial organs, and others. The advantage that the PHDA provides is bringing together expertise in machine learning with expertise in the medical sciences side and applying this in a translational setting. Pittsburgh is particularly primed for this type of work because we are the hotbed of these three intersecting areas – machine learning, 3D printing, and translational medicine.