Technically organ donation has become a much more successful process, but there are still considerable challenges facing the industry, both in terms of finding enough donors, but also finding donors that match.
Australian researchers have turned to the kind of AI algorithms that underpin many modern dating sites to try and improve matters and ensure a more accurate connection between organ donors and recipients.
“It’s a specially designed machine learning algorithm using multiple donor and recipient features to predict the outcome,” the team say.
The algorithm was trained using 25 distinct characteristics from both donor and recipient to understand what would happen to each organ graft. The algorithm considered things ranging from basic demographic information to the blood type and underlying disease of each person.
Once the algorithm was trained, it was able to accurately predict graft failure 30 days after the transplant with an accuracy of around 84%. This compares with accuracy levels of 68% using the current methods.
“It really meant for the first time we could assess an organ’s suitability in a quantitive way,” the team say, “as opposed to the current method, which really comes down to the position of the doctor eyeballing all the data and making a call based on their experience.”
Given the difficulties most countries have in securing a suitable supply of donors, ensuring those that are donated are successfully transferred is crucial.
The work has already been submitted for peer review, and the next stage is to undergo official randomized control trials to secure regulatory approval for the process. Whilst the early results are promising therefore, it is likely to be some time before we see the algorithm used in a live environment.
The team also believe that the data they collect around successful transplants can help to guide recruitment efforts to ensure that only the fittest donors offer up their organs, and only the fittest recipients request them.
It’s a fascinating project and one that should be watched with interest.