EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·23d ago·5 min read
Alibaba open-sources AI model that can detect cancer and nearly 150 conditions
TL;DR
Alibaba's Damo Academy has open-sourced Damo Radar, an AI model that detects nearly 150 abdominal conditions, including cancers, outperforming most radiologists in tests.
✦ Why It Matters
Explore integrating Damo Radar into diagnostic workflows to enhance cancer detection accuracy.
Key Takeaways
- Consider using Damo Radar for improved cancer detection in CT scans.
- Evaluate the model's performance against existing diagnostic tools.
- Explore potential applications in other medical imaging types.
- Stay updated on advancements in AI for healthcare diagnostics.
Full Summary
Damo Academy, part of Alibaba Group, has introduced Damo Radar, an advanced AI model designed to analyze contrast-enhanced computed tomography (CT) scans. This model can identify a wide range of abdominal conditions, including various cancers, by examining scans of 18 abdominal organs.
Trained on nearly 40,000 real-world examinations, Damo Radar achieved an impressive average area under the curve (AUC) of 0.913 across 146 clinical findings, where an AUC of 1.0 indicates perfect accuracy. The training involved pairing CT scans with clinical reports, enhancing the model's diagnostic capabilities.
Researchers believe that the methodology used for Damo Radar could be adapted for other types of medical imaging, positioning it as a pioneering generalist model in medical diagnostics.