DAMO LiON AI model finds 15 missed liver tumors in real-world trial

Alibaba's DAMO Academy, together with Shengjing Hospital of China Medical University, developed an AI model called DAMO LiON for liver cancer diagnosis. It finds tiny tumors in CT scans. In a two-month real-world trial, it caught 15 missed malignant tumors, mostly around 1 cm, helping patients get timely treatment. The study was published in Nature Medicine.

DAMO LiON works like an 'AI safety officer' helping doctors read scans. It spots tumors more accurately than radiologists, cutting reading time by 27% and boosting sensitivity by 11.5%. With AI help, junior doctors perform like senior ones. In two months, the AI reviewed CT scans from over 10,000 patients and helped find 15 missed liver metastases, changing their treatment plans.

DAMO Academy has worked on medical AI since 2017, with previous models for pancreatic, stomach, and colon cancer screening, also published in Nature Medicine. Now they're moving from screening to diagnosis, helping more patients get timely and accurate care.