Copy article

Liver Transplants: AI Now Gets Life-or-Death Decisions Right

ended 15. November 2025

The Lancer reports Stanford Medicine researchers prove the best use of AI isn't replacing human judgment, it's supporting it when the stakes are impossibly high.

Their machine learning model predicts donor death timing for liver transplants with 75% accuracy, outperforming surgeons' 65% rate. More critically, it reduced futile surgical preparations by 60%, saving resources, reducing healthcare worker burnout, and potentially getting more patients the organs they desperately need.

The system was personalised for clinicians. The researchers designed it to be customisable around surgeon preferences and hospital procedures. For example, the model can be set to calculate the time of death from when life support is removed or from when agonal breathing, a gasping breathing pattern that happens as a body is dying, begins. The researchers have also developed a natural language interface, similar to ChatGPT, that pulls information from the donor medical record into the model.

This is AI collaboration, not human replacement. AI handling pattern recognition across thousands of variables. Humans handling final decisions, ethical considerations, and patient care. Both getting better outcomes than either could achieve alone.

We want your views:

  • Would you trust AI to help decide if you or a loved one receives a life-saving organ transplant?
  • Should patients be told when AI is involved in their treatment decisions—and do they have the right to refuse?
  • If AI gets it wrong in a life-or-death situation, who's accountable—the doctor, the hospital, or the tech company?
  • Does this kind of AI free up doctors to spend more time with patients, or just pile on more pressure to process people faster?
  • Are we comfortable with AI having access to deeply personal medical records to make these predictions?
  • Should the NHS be investing in this kind of technology when waiting lists are already overwhelming staff?

1 responses from the Newspage community

Copy all

Star Quote
Copy

Colette MasonVerified
AI Ethics Consultant at Clever Clogs AI
Stanford Medicine researchers just proved what responsible AI actually looks like: quality training data, professional values respected, real-world clinical models, AI speed, and a human firmly in the loop. This is effective collaboration, not heartless replacement. AI does what it does best: detailed accurate pattern recognition across thousands of variables. Humans do what they do best: final decisions, ethical considerations, and patient care. Both get better outcomes than either could achieve alone.