A Stanford-led team developed a machine-learning model that predicts donor death timing, reducing canceled liver transplants and improving resource use.
Key Details
- 1A machine-learning model predicts whether donors will die within 30–45 minutes after life support removal, enabling timely liver transplantation.
- 2The model was trained on over 2,000 cases from six US transplant centers and outperformed surgeons’ judgment (75% vs 65% accuracy).
- 3Futile procurement rates (resources spent without a successful transplant) dropped by 60% using the model.
- 4Missed transplant opportunities remained similar to surgeon judgment (just over 15%), but further model refinements lowered this to ~10% in recent tests.
- 5The model incorporates donor demographics, vitals, lab results, ventilator settings, and neurological assessments, and features a natural language interface for clinical data extraction.
- 6It is customizable to clinical and hospital preferences and is being adapted for heart and lung transplantation.
Why It Matters
This research demonstrates the growing potential of machine learning in high-stakes surgical decision-making, supporting better allocation of scarce organs and healthcare resources. Increased accuracy and efficiency can enhance patient outcomes, reduce waste, and potentially expand AI's role into more imaging-driven and perioperative workflows.
Source
EurekAlert
Related News
•EurekAlert
FDA Approves Johns Hopkins AI Tool for Early Sepsis Detection
FDA clears an AI-driven system developed by Johns Hopkins to detect sepsis up to 48 hours earlier and reduce mortality rates.
•EurekAlert
AI-Driven Handheld Endomicroscope Enhances Early Cancer Detection
Researchers develop PrecisionView, a handheld AI-powered endomicroscope for real-time, high-resolution cancer diagnostics.
•EurekAlert
AI Model Uses EKG and EHR Data to Predict Sudden Cardiac Arrest
Researchers have developed AI models that analyze EKG and EHR data to predict risk of sudden cardiac arrest in the general population.