Alibaba's research division, DAMO Academy, publicly released a medical AI model capable of evaluating contrast-enhanced CT scans to identify various diseases and abnormalities, including cancerous tumors.
Named Damo Radar, the model is designed to examine scans covering 18 abdominal organs and identify 146 clinical findings, spanning nearly 150 medical conditions. According to the South China Morning Post, the development marks another major step in Alibaba's growing focus on artificial intelligence applications in healthcare and medical imaging.
ALSO READ: US Drafts Rules To Spare Pharma From Strictest China Investment Curbs: Report
Also Read | Layoffs Are Rising In Tech, But These Skills Are Helping People Unlock Better Pay
Across a test suite of nearly 40,000 real-world CT examinations, Damo Radar recorded an average Area Under the Curve (AUC) of 0.913 for 146 findings. In medical diagnostics, an AUC closer to 1 represents near-perfect accuracy in identifying conditions.
The model was also assessed in a study involving 26 radiologists from multiple hospitals. Researchers reported that Damo Radar outperformed 23 of the participating radiologists on average in the evaluation.
The study also examined how the technology could assist human specialists. When radiologists used Damo Radar as a diagnostic aid, the reported results showed a 10% reduction in missed diagnoses and a more than 30% reduction in diagnosis time.
The research was published in the journal Science. Alibaba has described Damo Radar as the “world's first expert-level generalist medical imaging model,” highlighting its ability to detect multiple conditions rather than focusing on a single disease or organ.
Researchers describe the system as "the world's first expert-level generalist medical imaging model," noting that its foundational training architecture can eventually be applied across various other medical imaging modalities.
ALSO READ: Trump Extends $100,000 H-1B Visa Rule Until 2027: What Changes For Indian Workers?
Expanding its medical AI portfolio, Damo Academy has turned its focus to early-detection screening tools for aortic dissections and pancreatic, stomach, and colorectal cancers.
Its April launch of the Coca AI model, developed in collaboration with Guangdong General Hospital and other domestic partners, demonstrated higher sensitivity than human radiologists in identifying early colorectal cancer on CT scans.
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.
