Doctors have developed an AI tool that can reduce wasted effort in organ transplants by 60%.
Thousands of patients around the world are waiting for potentially life-saving donors, and there are more candidates on the waiting list than organs available.
Recently, when liver transplantation is required, access has been expanded through the use of donors who have died after cardiac arrest. However, in approximately half of donations after cardiovascular death (DCD) cases, the transplant is ultimately canceled.
That’s because the time from when life support is removed until death cannot exceed 45 minutes. Surgeons often refuse a liver transplant if the donor does not die within the time required to preserve organ quality, increasing the risk of complications for the recipient.
Now, doctors, scientists, and researchers at Stanford University have developed a machine learning model that predicts whether a donor is likely to die before the organ can be transplanted.
This AI tool outperformed leading surgeons and reduced by 60% the rate of wasted procurements that occur when transplant preparation begins but the donor dies too late.
“By identifying when an organ is likely to be useful before starting preparation for surgery, this model could make the transplant process more efficient,” said Dr. Kazunari Sasaki, clinical professor of abdominal transplantation and senior author of the study.
“It also has the potential to make organ transplants available to more candidates who need them.”
Details of this breakthrough were published in the Lancet Digital Health journal.
This advancement could reduce the number of cases in which organs are prepared by health care workers for recovery but deemed unsuitable for recovery or transplantation, placing financial and operational burdens on transplant centers.
Hospitals estimate this critical period primarily based on surgeon judgment, which can vary widely and can lead to unnecessary costs and wasted resources.
The new AI tool was trained on data from more than 2,000 donors across multiple U.S. transplant centers. Uses neurological, respiratory, and cardiovascular data to predict progression to death in potential donors with greater accuracy than previous models or human experts.
After newsletter promotion
The model was tested retrospectively and prospectively and was able to reduce wasted procurement by 60% compared to surgeon predictions. Importantly, the researchers said, accuracy remained even when some donor information was missing.
Reliable, data-driven tools help medical staff make better decisions, optimize organ use, and reduce wasted effort and costs.
This approach could be an important step forward in transplantation, the researchers said, highlighting the “potential for advanced AI techniques to optimize organ utilization from DCD donors.”
Next, they plan to modify the AI tool and try it on heart and lung transplants.
