Alibaba open-sources medical AI model that can detect cancer and nearly 150 conditions

Alibaba Group Holding’s research division, Damo Academy, has released an open-source artificial intelligence model named Damo Radar. This advanced vision-language model is capable of detecting nearly 150 medical conditions, including various cancers, by analyzing computed tomography (CT) scans of 18 abdominal organs. The model was trained using CT scans and corresponding clinical reports, ultimately achieving an impressive average area under the curve (AUC) of 0.913 across 146 clinical findings in over 40,000 real-world examinations. An AUC of 1.0 denotes perfect diagnostic accuracy.

In a comparative study involving 26 human radiologists from multiple hospitals, Damo Radar's average accuracy outperformed 23 participants. The model also helped radiologists improve their ability to prevent missed diagnoses by 10% while reducing the time required by more than 30%. Damo Academy published the research in the journal Science, with contributions from several institutions, including a hospital affiliated with Zhejiang University.

This release follows a series of medical AI advancements by Alibaba, as Chinese technology companies increasingly leverage cutting-edge technology in healthcare settings. Recently, Damo Academy developed the Coca AI model, which reportedly detects early-stage colorectal cancer more accurately than radiologists using CT scans. The model was co-developed with Chinese institutions, including Guangdong General Hospital.