Artificial intelligence applied to ultrasound scans can predict a baby's delivery date with high accuracy. The AI model, developed by Ultrasound AI, analyses ultrasound images to estimate the time until delivery. Testing has shown the AI's accuracy improves with retraining, achieving an accuracy measure of 0.95 for term births.
The AI was trained using over two million ultrasound images from women who delivered at the University of Kentucky. The AI's performance remains consistent across all trimesters and patient demographics. This technology could be especially useful in areas where access to specialist care is limited, enabling early risk identification.
The AI's ability to learn sequentially allows it to better predict preterm births, potentially leading to further advancements in the future. Experts suggest this technology marks a significant advancement in maternal-foetal medicine, offering a non-invasive method to improve pregnancy outcomes.
Subscribe for Weekly Updates
Stay ahead with our weekly AI and tech briefings, delivered every Tuesday.
