It was January and the streets were covered in thick layers of ice. One chilly morning, while shoveling snow in my front yard, I have managed to successfully fracture a bone in one of my fingers. I could immediately envision the struggles of being bound in plaster for the coming weeks: trying to bathe with my arm covered in plastic bags, the constant itching and the everyday nuisance of getting dressed properly. On my way to the emergency department, I was wondering if it would be possible to get rid of the annoyances of traditional plaster casts by creating a weblike 3D printed cast?
Nowadays all the necessary technology is readily available for the task and with some holistic thinking one would be able to design and manufacture a 3D printed cast DIY style.
Sitting at the A&E
I was still hoping that I would get off with a simple bone bruise, but when the X-ray came back, it made it clear that there is a fracture in my right ringfinger. Luckily, the bones have not shifted, but the doctor told me I would have a nice cast for the coming weeks. While I was waiting for the cast to be applied, I had enough time to think about how modern technology could help me in this situation. I have browsed online for possible solutions and started connecting the dots. But what about the doctor? What will he say about an idea that conforms to none of the current health specifications and standards?
Disclaimer – DO NOT do it yourself
I knew that I was risking the proper healing of my hand with a self made cast, as it was impossible to acquire all the required permits for conducting my experiment under the supervision of a medical professional.
Scanning the hand
The first step was to create a triangulated surface of my hand, preferably in STL format. I knew many companies who are able to do that with the aid of a portable laser scanner, but I needed one who also have experience in scanning human bodies in an anatomically correct way. After a quick search, I have decided to schedule a meeting with the Basiliskus3D team, who have a great expertise in scanning people.
In their office we have selected the best scanning method for this particular task, the 3D white light scanning. We have selected this method, because it enabled us fast and accurate measurement of organic surfaces without the need for markers (orienting points on the surface of the object).
It was important for us to be able to scan the fractured arm from all sides with multiple sweeps (subsequential interrupted scans) and the software of our chosen scanner had to be able to properly join the seperate scans. In our case, the fractured limb would have had to be held according to the proper casting position, with my elbow on the table. However, it was not. This was one of my first mistakes from a medical point of view.
The scanning process was quick and painless. It only took 3-4 sweeps to construct a triangulated STL model of my hand. The STL file was assembled with the software of the scanner and the company sent it for me in a few days. Meanwhile, I had time to think about the design of the cast itself…
Designing the cast
I have found some active projects I could borrow ideas from to develop my basic concept, but none of them suited my purposes. I have designed the cast for my specific needs and not for general usage, and this is also reflected in the design aspects.
- Split halves – I had to find a solution that makes it possible to remove and attach the cast easily for post production modifications. I have decided that the cast would comprise of two separate sides held together by rubberband fixtures, not via snap fit clipping.
- Reduced size – Securing my fracture in position would have required a full arm cast. However, due to production limitations I had to reduce the size.
- Altered immobilization area – Normally, one would have to immobilize all fingers within the cast, but I wanted to leave the index finger free for using the mouse, so I could keep on working.
Design workflow, problems and software
After the scans gave me the triangle model, I had to choose between two solutions. I could either cover the triangle model with surfaces (with a certain precision), and create the cast by manipulating the solid model. The downside of this method is that it would take a long time to create a solid body with the desired accuracy. Another solution is doing the design work directly on the triangle model. However, this would prove inconvenient as there are many steps during the design process which would be cumbersome to do this way.
My final solution was a hybrid method, which was possible due to a feature of the Autodesk PowerShape software. In this CAD system we can easily manipulate surfaces, solid bodies and triangle meshes and intersect them freely.
I have created a solid body which was an approximation of the final cast form and during the last phase of the design process, I have extracted the slightly offseted triangle model of my hand, ensuring proper fit. Afterwards, I had to define the split position. The process was the same as the method used in plastic surface splitting in manufacturing.
The next step was to design a hollow mesh for the solid tube of the cast to reduce weight and provide ventilation. During weight reduction I had to maintain proper strength and stiffness at the key fix points, so the final cast would still hold my hand in place. All methods for putting holes in the solid material would be suitable for our purposes, provided that they take these aspects into account. For example, we could customize the cast by cutting specific patterns or text in the material.
I chose a cell structure pattern, the Voronoi splines, which can be also found in nature. In case of 2D solids, we can create them by bisecting the line connecting two neighbouring points, then repeating the process for all of our points. After intersecting the bisectors at their nearest crossing point, we create splines within the closed areas. We can try this in action with the Interactive Voroni Diagram application.
Of course, there are several programs for generating Voronoi meshes. For example, Voronator processes our 3Ds model within minutes. There is also a free Voronoi plugin for Fusion 360, but neither of these options had the feature of manually adjusting the size and spacing of the cells on a solid. I managed to solve this problem with an obscure little graphics program, called Inkscape. This free software had an option for creating Voronoi diagrams with manually selectable points, enabling me to adjust the mesh density of the pattern.
Now I just had to figure out how to distribute the 2D mesh on a solid. The solution was the 2D curve wrapping and unwrapping function of Powershape. Unwrapping the outer contour of the surface provided me with a shape, inside of which the mesh would be generated. Afterwards, the curves had to be wrapped around the surface and the cuts had to be made on the solid.
After hollowing out the solid, I have designed locking elements on the two halves, which would be held together by rubberbands. The locking elements had smooth curves so they would not get caught in clothes and alike.
Manipulating the triangle mesh
The finished model was in triangle (STL) format, as I had to extract the scanned model of my hand from the solid of the cast, in order to avoid complete reverse engineering. However, it can be difficult to do processes such as rounding the edges on triangle models. For this reason, I have utilized the help of Autodesk’s free software for triangle mesh manipulation, Meshmixer, which allowed me to manipulate STL models with ease, not unlike a 3D Photoshop. I have used this software for smoothing the triangle models, including the scan of my hand, rounding the edges and preparing the models for 3D printing.
3D printing the cast
The most difficult part of my venture was 3D printing the cast, since the machine had to be of considerable size to be able to print out the parts. Accuracy did not concern me, considering the lack of precision in traditional immobilization methods. Most commercially available 3D printers would have enough accuracy for these purposes. In the end, the cast was printed by Varinex Zrt., a Hungarian manufacturing company (for whom I owe a big thank you for their help).
Before the printing process I have sent them the cast models in STL format, so that they would be able to choose the proper technology and machine for this task. Varinex have chosen a Stratasys Fortus 400mc Large FDM technology machine with 0,2mm accuracy. We had to take into account that the printing material would get in contact with my skin for a prolonged period and it would need to be strong enough to sustain my broken arm. The biocompatible PC-ISO material fulfilled all of our requirements. This material is regularly used for manufacturing medical devices and food packaging. The supports that were necessary to print the special shape of the cast were automatically defined by the printer’s software and printed with a cheaper material with less density. After the printing process, the cast was removed from these supports. It took 10 hours and 150ccm of raw material to print out the final cast.
Test and conclusion
After the cast was finished I was glad to ditch the heavy and uncomfortable plaster for a modern and far more pleasant experience. It was time to see how this brand new immobilization device would fare in a real world test.
During the 3 week long live test I have noticed several advantages of the new technology. My hand was well ventilated, it was not drenched in sweat and I did not have to wrap myself in plastic bags just to take a bath. I could clean my arm properly, dress with ease, and it was light and comfortable to wear. But instead of listing the advantages I would rather try to answer the most important question: did the 3D printed cast have adequate healing properties?
During the real life test I have came across a few problems, as usual with prototypes:
- during the scanning process I have failed to hold my arm in the proper position, and I had to modify the cast afterwards to make my pinky fit snugly
- my arm was still swollen during the scanning and the final cast became a bit loose
An X-ray provided the evidence for the success of the cast: the fracture healed properly! This concluded the real world test on a high note. It has been confirmed that traditional methods can be adequately replaced by newer technologies in the future. Of course, the described method is not suitable for a wide range of fractures, as there are many more aspects to consider when dealing with fractures of the leg and other large bones, and open fractures require completely different immobilization techniques. However, 3D printed casts could offer a better and more comfortable alternative to traditional methods in many situations.
Project by: József Kollár