While Subburaj Karupppasamy was completing his university degree, his brother in India suffered a spinal injury in an accident. He went in and out of hospitals for a long time until he finally had a spine surgery with implants put in.

Both the operations and implants were very expensive. Fortunately his brother is a businessman and could afford it, but many in India would not have been able to. This experience led Subburaj to realize the potential for developing more affordable implants using computer modeling

- I was studying medical engineering at the time, and I was interested in shape complexity. The most complex object to study is the human body and when my brother had his accident, I thought that studying spinal injuries and the technology to fix them was what I needed to do, he says.

Subburaj Karupppasamy embarked on a mission to develop $2,000 spinal implants. At the time a spinal implant was about $10,000. A goal he’s still trying to achieve.

Today he leads a group of scientists at the Department of Mechanical and Production Engineering at Aarhus University. Through advanced computer models the group is trying to develop new implants. Not only for spinal surgery, but for a number of different injuries.

- We want to use our computer models to develop the next generation of implants. With more precise data we can make implants that are not only more functional, but also cheaper and less intrusive, he says.
 

A model of both spine and muscles

One of the most common spinal diseases is called scoliosis. Two to three percent of the population is suffering from it causing their spine to curve sideways.

If the disease is not treated it often results in chronic debilitating back pain. But if scoliosis is discovered in early childhood, it’s possible to correct the curvature of the spine. Today that is done with implanting titanium rods through the spine.

According to Subburaj Karupppasamy that treatment is both expensive and prone to failure.

- The traditional treatment has a failure rate of around 60 percent. Just think about that number. In more than half of the cases the treatment, which is very intrusive, fails.

- One of our research projects aims to develop an alternative to this treatment. We use computer models of the spine and the attached muscles to determine the optimal placement for the implants.

One of the reasons that scoliosis treatment often fails, is that the rods have been attached to the wrong places. This sometimes results in the rods snapping loose or the ribs breaking – and that leads to the need for further surgery.

- Scoliosis patients often have surgery multiple times to put the rods back into place. A couple of years ago a company in San Diego came up with a new approach. Instead of undergoing surgery again the rods are put back into place with a scanner using magnetic force. Unfortunately the failure rate is also high here.

- Right now we are trying to redesign these rods and instead use pistons. We are doing calculations that show us exactly how to design them and where they need to be anchored. When we have the optimal solution we will 3D-print them and cooperate with doctors in clinical trials at AUH and in Munich.

Technology ready for the ER

Subburaj Karupppasamy’s team is not only working on implants. They have developed a computer model that can predict the risk of a patient having fractures in the spine. A technology that is almost ready to be implemented in the hospitals.

- We accessed CT scans of nearly 400 patients who were scanned for reasons other than spinal problems. In this data we looked for spinal fractures and from the fractures we found, we created a risk score model.

- Many patients – especially elderly people – don’t know that their bones are weak. They never discover that they have minor fractures on the spine. But the second fracture often disables them and ties them to a wheelchair. Our model can calculate the patient’s risk of having fractures pretty accurately.

Subburaj Karupppasamy hopes that the hospitals will embrace their new technology, because today these fractures often go undiscovered.

- Today hospitals use something called DEXA-data. But it’s prediction rate is pretty low. It’s only correct about 50 percent of the time. We can give a much better risk profile.

- In the future I hope that our technology will be standard. When people are screened for breast cancer at age 40, maybe we could also scan for spinal fractures. The degeneration of the bones often start around the same age when women go into menopause.

Best of both worlds

What sets Subburaj Karupppasamy’s research group apart is that they combine the language and knowledge of two very different worlds: The engineering and the medical world. These are two research areas that typically don’t intersect, explains Subburaj Karupppasamy.

- Most of the medical research is focused on new types of medicine, not mechanical solutions. The money is in developing new drugs. This means that most of the technology used for implants today is pretty low tech.

The implants and treatments for some diseases haven’t been updated in decades. They are, as Subburaj Karupppasamy says, based on very old technology. But new and powerful computers, AI and 3D-printing have made it possible to develop all sorts of new implants.

- In our view there’s room for huge improvements and for making a significant impact in people’s day-to-day lives. We have all these new computer tools but we haven’t really used them that much in this area of research. That is what we aim to do.