3D-printed model brains are more accessible and ethical than conventional alternatives, says CU Boulder researcher Jennifer Stratford

When studying anatomy, students often use models, which recreate the structure of an organ or other part of the body in a form that is easy to study. 

Although donor bodies are generally considered the gold standard in higher education because they are the only resource that allows students to directly study real anatomy, the number of donor bodies is limited, making them difficult to access, especially in low- and middle-income countries.

Models are somewhat more accessible, but if they are anatomically accurate, similar challenges apply, because they must be based on real donor anatomy to provide educational value. Accurate models are also expensive. However, if educators could 3D print their own models, cost would be less of a barrier, says Jennifer Stratford, a University of Colorado Boulder teaching professor of psychology and neuroscience.

Jennifer Stratford is a CU Boulder teaching professor of psychology and neuroscience.

In a recent analysis conducted with colleagues in the CU Anschutz Modern Human Anatomy Program as well as other institutions, Stratford details the learning benefits of 3D printing brain models. 

Why brain models?

Neuroanatomy, or the anatomy of the brain and other parts of the nervous system, is a challenging yet important area of study, according to Stratford. Clinical specialties like neurology, psychiatry, anesthesiology, radiology, physiotherapy and more rely in part on knowledge of the brain to correctly diagnose and treat patients. However, Stratford explains, “the human brain is much more difficult to understand than other organs.

“If I showed you a model heart, you could make some educated guesses about how the heart works,” she continues. On the other hand, “if I give you a model of a brain, and I asked you how you think it works just by looking at it, it is almost impossible to guess. That makes learning neuroanatomy tricky,” Stratford adds, and it makes resources that can help students learn particularly valuable.

“I think a lot of people do not realize all the effort, expense and ethical considerations that go into a human brain model,” Stratford says. “It can cost hundreds—if not thousands—of dollars to buy that model from a company.” Additionally, “a lot of original brain models were created without asking for permission,” whether from cadavers that a lab had access to or from bodies that were donated without specific consent to make models based on the brain.

While brain models are more accessible than donor brains, they can still be difficult to obtain, especially in large enough numbers for a whole class of students. 3D printing addresses this by giving educators the option to produce anatomical models themselves at a reduced cost, either using brain scans they take themselves or data from open-access repositories like the CU Modern Human Anatomy Hub, which was created by the research team behind the paper. Putting scan data online comes with additional ethical considerations, but the authors provide ways of managing these, such as data anonymization, and online resources make it possible for people all over the world to create their own models.

“What 3D models help us do is to give students, other educators and the public the opportunity to hold some of these specimens in their hands,” Stratford says.

Collecting and processing data

Stratford and her co-authors describe a five-step process that educators can follow to create and use a 3D model, beginning with selecting or acquiring a model of the brain. These models are ready to use, but if an educator wants a custom model, it can be created from patient data or a physical, donated specimen.

A revised Bloom's taxonomy applied to 3D prints of neuroanatomy, explaining learning objectives at each step of the taxonomy. (Infographic courtesy the paper authors)

Patient data includes CT and MRI scans of the brain or other nervous-system structures. A 3D object can be extracted from this data using a process called segmentation. 

“All the patient data, at least that we use, is digitally created with a computer, but it does not have all the brain structures picked out,” Stratford says. “It is not very clean, so one of the goals of extraction is to clean up the 3D rendering of the brain, both making it more crisp, but also highlighting different areas.” 

This process also can be described as “removing artifacts.” In this context, artifacts are errors introduced by the way information is captured, such as a lens flare in a photograph.

Additionally, the different regions of the brain are not easily distinguishable from raw data, so highlighting some of them can make the final product a better teaching tool. According to Stratford, while some programs can segment data automatically, they are still new and error prone, so most people currently do segmentation by hand, and it can be one of the more labor-intensive parts of the rendering process.

Physical specimens can also be turned into 3D models using scanning technology, including commercial 3D scanners or even digital cameras or smartphones. Scanners make models by measuring the distance between the scanner and the surface of the object, while cameras algorithmically combine a series of photos in a process called photogrammetry. Scanning also introduces some artifacts, so digital editing is necessary either way, and it is considered the second step in the process.

Thirdly, Stratford and her co-authors recommend licensing the model with a Creative Commons license and making it open access online. Creative Commons is a nonprofit organization that has developed free copyright licenses designed to help content owners reserve some rights while waiving others. In this case, a Creative Commons license would allow the owner or owners to retain control over how their work is used while making it free to access. According to open access principles, research literature and especially medical resources should be free so that healthcare providers are not prevented from taking advantage of potentially vital information by paywalls or other such barriers.

 

An example of 3D brain prints by printing level. (Infographic courtesy of the paper authors)

Printing and using models

The fourth step is to actually 3D print the model. This requires converting data collected in the first step into instructions the printer can use, which is done by software applications called “slicers,” referencing how most 3D printers create physical objects by laying down material in layers or “slices” of the object. 

“That’s what we do in 3D printing in general,” Stratford says, “not just for brains. It is like you are printing multiple sheets every time, and each sheet is made of little bumps in different places, and then you push those together to make a 3D rendering.”

This step can vary significantly in terms of cost and complexity depending on the model of printer and type of materials used. To reflect this, Stratford and her colleagues split the printing process into three difficulty levels ranging from beginner to advanced. According to Stratford, while the beginner level is cheapest and requires only a basic 3D printer, it limits the product to a single color.

 “They are really cheap and easy to make, so those are the type I typically use when I do outreach. If somebody drops something, I am not worried if they get damaged, and I am not looking for a ton of detail there.

“From there, you can get more sophisticated. Different 3D printer models that are more expensive can print, for instance, in two colors,” she continues, allowing different parts of the brain to be color-coded. “For instance, you could highlight what is called the spinothalamic tract. That is one of the classic pathways in the brain, but anatomically, it is hard to pick out on its own.” 

Multiple colors help with this by creating contrast. The most expensive 3D printers allow for the use of more than two colors. 

“One of the things we used to 3D print” at Anschutz, Stratford says, “was taste buds from your tongue. In your taste buds, there are lots of different cell types, and so we could make each cell type a different color and then have that printed in the 3D model. It was really cool to see how that looked.”

Besides different colors, printers can use different filaments (the materials the models are made from), some of which are easier to work with than others. A beginner print may use bioplastic (PLA) filament, which is less expensive and does not have a strong odor, but models made from PLA can be brittle. Advanced printing may involve multiple materials or higher-quality materials like resin. However, both require more expensive printers, and liquid resin releases toxic fumes, so personal protective equipment is needed when working with it.

Finally, once the model has been printed, it can be used for education. 

“One of the things they are really useful for,” Stratford explains, “is when you want to teach the overall structure of the brain. That is called gross neuroanatomy. Being able to give individual groups of students their own models can be helpful. Because traditional models are so expensive, most classrooms cannot afford to have more than maybe two or three.” 

Because 3D-printed models are cheaper, the ratio of models to students can be better. “For instance, in my class of about 45, we had 15 models to work with, so even though people had to triple up, it is much better than having only one or two models for people to look through.

“Another aspect is, you are generating models based on real brains, whereas if you buy an anatomical model from a company, that is more like a cartoon of the brain. So, what is nice is that you can show a student what your brain actually looks like. Students really do identify with that. I will ask students, ‘What is the biggest thing that sticks out in your mind from this semester?’ And many of them will say it is the 3D models, because it seems so personal.”

For example, models can show what different neurological diseases and types of brain damage look like as well as how they differ from each other.

“It addresses so many different needs. It is better from a teaching perspective, the ethics are better compared to past options, and it is just more widely available,” Stratford says. “Growing up and as a young person, I did not have a lot of access to these things, just because they were so expensive. So, I am really excited about the opportunities this provides to people.”

James P. Lister, Andrew Cale, Zach D. Stetter, Jamie E. Cronin, Steven R. Summers, Paul G. McMenamin and Maureen E. Stabio are co-authors of this paper.

Did you enjoy this article? Subscribe to our newsletter. Passionate about psychology and neuroscience? Show your support.