Alibaba releases RADAR, expert-level AI for abdominal CT diagnosis
Alibaba DAMO Academy open-sources a vision-language model trained on 400,000+ CT scans for radiology interpretation.
What to know
- Alibaba DAMO Academy released RADAR, an AI model trained on 400,000+ abdominal CT scans to match expert radiologist performance.
- The model uses 15 million image–text pairs derived directly from clinical reports, eliminating need for manual annotation.
- RADAR is available open-source under third-party-licensed components, providing a framework for scalable radiology AI development.
“RADAR is a generalist vision-language model trained on over 400,000 contrast-enhanced abdominal CT examinations with 15 million anatomy-aware image–text pairs, learning directly from clinical reports without manual annotation.”
Alibaba DAMO Academy · GitHub repository ↗
Alibaba DAMO Academy Model developer and publisher
How it unfolded 1 development · click the chart to see its coverage articlesposts
-
1
Alibaba DAMO Academy releases RADAR open-source model
Alibaba DAMO Academy published RADAR, a generalist vision-language model trained on over 400,000 contrast-enhanced abdominal CT examinations with 15 million anatomy-aware image–text pairs. The model learns directly from clinical reports without manual annotation and demonstrates expert-level performance on routine and complex clinical tasks.
-
first by Alibaba DAMO Academy on GitHub, 6d ago
-