Medical professionals have created an AI tool capable of decreasing wasted efforts in organ transplants by 60%.

Across the globe, thousands of patients await potentially life-saving organ donations, with more individuals on the waiting list than available organs.

Recently, the scope of liver transplants has broadened to include donors who have passed away from cardiac arrest. However, in around half of the cases involving donations after cardiovascular death (DCD), the transplant is ultimately called off.

This occurs because the duration from the removal of life support to the moment of death must not exceed 45 minutes. Surgeons frequently decline to proceed with a liver transplant if the donor does not pass away within the timeframe necessary to maintain organ viability, which increases complications for recipients.

Now, a team of doctors, scientists, and researchers at Stanford University has developed a machine learning model that forecasts whether a donor is likely to pass away before the organ can be transplanted.

This AI tool has surpassed leading surgeons, cutting down the rate of wasted procurements—where preparation for a transplant begins but the donor dies too late—by 60%.

“By pinpointing when an organ is likely to be viable before initiating surgical preparations, this model could enhance the efficiency of the transplant process,” stated Dr. Kazunari Sasaki, a clinical professor of abdominal transplantation and the study’s senior author.

“It also has the capability to make organ transplants accessible to a greater number of candidates in need.”

Here are the specifics of this breakthrough: Published in Lancet Digital Health journal.

This advancement could lessen the instances in which organs are prepared for recovery by healthcare workers but are deemed unsuitable for transplantation, imposing financial and operational challenges on transplant centers.

Hospitals primarily estimate this critical period based on the judgment of the surgeons, which varies significantly and can result in unnecessary expenses and wasted resources.

The new AI tool was trained with data from over 2,000 donors from various U.S. transplant centers. It utilizes neurological, respiratory, and cardiovascular data to predict the likelihood of death in potential donors with greater accuracy than previous models or human specialists.

The model was tested both retrospectively and prospectively, successfully reducing procurement waste by 60% compared to surgeon assessments. Notably, the researchers indicated that accuracy was upheld even with some missing donor information.

Reliable, data-driven tools assist medical professionals in making informed decisions, optimizing organ usage, and minimizing wasted efforts and costs.

This method could represent a significant advancement in transplantation, the researchers emphasized, showcasing the “potential for advanced AI techniques to maximize organ utilization from DCD donors.”

In the next phase, they plan to refine the AI tool and test it for heart and lung transplants.