Missouri S&T professor is looking to AI to help with kidney transplants
A Missouri University of Science and Technology researcher is looking for ways artificial intelligence can improve how kidney donors are matched with recipients.
Missouri S&T engineering management assistant professor Casey Canfield is leading a research effort to develop AI software that can streamline the matching process.
Her preliminary research shows that excellent kidneys find their way into people on the top of the transplant list efficiently, but Canfield believes there is room to improve placing less-than-perfect kidneys.
“What we really need to find is the person in the middle of the list who would really benefit from getting a transplant sooner rather than having to wait two-three years to get a transplant,” Canfield said.
Vetting kidney donors and people on the transplant list is a thorough process but can slow down as less desirable organs start at the top of the list and work their way down.
A potential recipient at the top of the list is likely to wait for a better match.
When time is running out and there is a kidney available that might be right for someone in the middle of the list, doctors can start a process called “accelerated placement” — which allows doctors to bypass some of the usual steps in placing a kidney.
“To date, there hasn’t been a super systematic way of figuring out when a kidney is hard to place, when do you pull the trigger and decide it’s time to switch into accelerated placement,” Canfield said.
That’s where software powered by AI could help. Canfield is quick to say the software would not make medical decisions or even recommendations to doctors, but rather give them options and more information more quickly.
Canfield is working with SSM Health Saint Louis University Hospital on the four-year study. It received almost $2 million in funding from the National Science Foundation.
In addition to improving the speed of matching donated kidneys with patients, Canfield said, the AI could identify any unintentional bias that might exist in the current system.
“Right now we are in the phase where we are trying to measure and analyze, for example in terms of race, what are the outcomes that we are seeing? Is our algorithm making the outcomes more fair, less fair, and what do we even mean by fairness?” Canfield said.
Canfield is working with SSM Health St. Louis University Hospital, United Network for Organ Sharing and MidAmerica Transplant, a local organ procurement organization on the study.
Dr. Henry Randall, professor of surgery at St. Louis University School of Medicine, believes the AI decision support system that will be developed will play a critical role in improving the lives of those waiting for kidneys.
“[This project] will help surgeons and organ procurement organizations work together to make data-driven decisions on organ acceptance,” Randall said.