A Parametric Human Spine That Could Simulate Scoliosis for Medicine




Last month, Ioan Valentin Dascalescu, a first-class honors student at ITT-Dublin (Institute of Technology Tallaght), delivered a report on a CAD design project he hoped would help surgeons better treat scoliosis.

Scoliosis is a curvature of the spine that typically develops during childhood and can lead to disability. In his report, Dascalescu presents a 3D parametric model that simulates the disorder.

But Dascalescu’s  model isn’t just accurate and sophisticated. He actually designed it so that anybody can use his work to quickly generate a new spine model that reflects the anatomy of individual scoliosis subjects. That is, by changing the parameters, surgeons can closely replicate the spines of actual patients.

His approach? A  bi-directional model using Creo and PTC Mathcad.

Image: Dascalescu spine assemblies adjusted for different stature and curvature.

Anatomy lesson

Unless you’re good friends with your chiropractor, you might need a short Anatomy 101 crash course before we dive into the details of Dascalescu’s model.

Here’s what you need to know about your backbone:

  • The spine is divided into five major regions—Cervical, Thoracic, Lumbar, Sacral, and Coccygeal.
  • The Cobb angle is a measurement used to determine the severity of scoliosis. In the model shown in the figure above, the Cobb angle of the spine is 30 degrees.

Dascalescu says, “Scoliosis and its degrees of severity are best described in terms of the Cobb angle. During my research, I found that there is a height loss associated with spinal curvatures, which had to be taken in consideration when calculating the final spine height that was parameterized. This was essential for the validation of the model.”

The design challenge

Of course, Dascalescu isn’t the first to model a spine, but he found previously published spines highly simplified. “Early models were very crude,” writes Dascalescu in his report. “They used simple shapes such as circles and ellipses to create the geometry, with only one master vertebra to model the whole spine. “

Dascalescu thought he could do better, especially with the technology available today.

Image: A previously published spine model (left) and Dascalescu’s model (right).

So, where does someone start to design a better spine? To narrow the scope of his project, Dascalescu focused his efforts on developing a parametric model of only two of the five spinal regions—the cervical and thoracic regions.

He used validated STL file models of the vertebrae to create master vertebrae for cervical and thoracic regions. Next, he collected data from an anthropometric database and scientific papers to parameterize the vertebrae and establish relationships between the parameters.

Image: Evolutionary pictures of Dascalescu’s work assembling the thoracic region of the spine.

“Building the spine spline was particularly demanding, as the published data did not present a viable equation that could allow the spine to be parameterized in the various planes that surgeons look at,” says Dascalescu.

“The Creo-Mathcad interface required particular attention when setting up the inputs and the outputs”  To make the work even more challenging, he limited his assembly to only two input parameters, stature and Cobb angle. With just two parameters, he could make it easier for the users who would later need to customize the model.

Creating a bi-directional interface

From there, he created a bi-directional framework system between Creo and Mathcad that will allow data transfer in an effective manner between the two packages. Here’s how it works:

Cobb angle and stature parameters are adjusted in Creo . The new parameter values are sent from Creo to PTC Mathcad, which then calculates all of the new dimensions based on the two parameters. Then, the data gets sent back to Creo and the model is regenerated. Height loss due to the Cobb angle was also taken into consideration.

In short, changes anyone makes to the parameters in Creo automatically get sent to PTC Mathcad, formulas are run, and then changes are sent back to Creo and the model is regenerated.

Key takeaways

“Parameterizing a whole geometric model in terms of a couple of parameters can save huge amounts of time and money,” Dascalescu says. “Once the geometry is captured, with the help of some relatively simple relations, a model can be regenerated in minutes instead of hours.”

For Dascalescu, designing a whole part in one sketch wouldn’t have worked well. But by can breaking it down into smaller parts, extruding  those asymmetrical into solids, and then joining them together he was more effectively able to create a complex model.

His advice to others?  “Simplify the model and test it as you go along. When the modelling principle is established, then it can be scaled up in size and complexity.”

Check out Dascalescu’s full report here.

What can you do with Creo and PTC Mathcad together?

If you think you might be able to make the world a better place too by integrating engineering calculation software with your 3D design software, explore PTC Mathcad.  The helps you perform, analyze, document, and share your calculations easily. And here’s the best part—there’s a free version you can download today.

Cris Forster builds one-of-a-kind musical instruments with unconventional intervals using PTC Mathcad