More than 100 million people worldwide require a urinary catheter every year. The devices can be lifesaving, especially after surgery. But many of those who used them—about one quarter in developing countries and about an eighth in the U.S.—may develop a catheter-associated urinary tract infection (CAUTI), commonly caused by bacteria building up inside the tube.
Now, with help from artificial intelligence, researchers have designed a new catheter that they say could reduce bacterial contamination by up to two orders of magnitude—without antibiotics. Its interior is studded with three-dimensional geometric shapes that help prevent bacteria from gaining traction. “These are pretty exciting results,” says Glenn Werneburg, a Cleveland Clinic urologist, who was not involved in the new study, which was published recently in Science Advances.
“In a normal catheter, there is no physical shape inside,” says study co-author Animashree Anandkumar, a computer scientist at the California Institute of Technology. This leaves a smooth highway for bacteria to work their way up from the outside and colonize the inner surface. When colonies build up in the catheter, near the bladder, they can enter the urinary tract, leading to a CAUTI.
In the past doctors have sometimes coated catheters’ interior walls with antibiotic drugs or metallic agents such as silver to kill bacteria. But such methods can be expensive—and also ineffective as antibiotic-resistant bacteria become more prevalent. The new device does not rely on a specialized coating to repel microbes; simple geometry does the trick. A series of tiny, 3-D-printed ridges that are shaped like sharp triangles line the inside of the catheter, forming a kind of obstacle course for the bacteria. As the microbes try to swim upstream, they bump and tumble into the ridges, eventually halting or bouncing downstream. The design could help cut down on expensive and unnecessary antibiotics and could prolong the amount of time a catheter can be used.
To find the perfect bacteria-repelling labyrinth, Anandkumar and her team used AI to quickly run digitally modeled catheters through tens of thousands of simulations. Once they landed on a design that best blocked virtual bacteria under multiple scenarios in the computer model, they 3-D printed a prototype and tested it in the lab with a broth containing Escherichia coli bacteria. After 24 hours, the experimental device had built up nearly 100 times fewer bacterial colonies than a traditional catheter that the researchers 3-D printed and tested alongside it.
The new catheter is currently only optimized to resist E. coli—one of the most common microbes associated with CAUTIs. Other species are also known to colonize catheters and cause infection, however. “Bacteria on catheters are present as biofilms, and we know that different species of bacteria behave in different ways,” Werneburg says. He adds that the team may need to develop a modified future design that is also inaccessible to other microbes, such as Enterococcus and Proteus bacteria.
Anandkumar agrees, and she says that accurately modeling new designs may require further research—and more data about these microbes’ physical and chemical properties. The scientists will also need to test their design in a clinical setting before it can be widely produced. And Anandkumar says AI modeling has potential far beyond catheters: she hopes to harness AI to help design drugs, energy-efficient airplane propellers, and more. “To me,” she says, “this is just the beginning.”