Alibaba open-sources Damo Radar AI for abdominal disease detection

Alibaba’s research division, Damo Academy, has open-sourced Damo Radar, a vision-language artificial intelligence model designed to identify approximately 146 to 150 clinical findings from abdominal CT scans. The model analyzes contrast-enhanced examinations across 18 abdominal organs to detect various diseases, including malignant tumors and other abnormalities.

In evaluations involving nearly 40,000 real-world examinations, the system achieved an average area under the curve (AUC) of 0.913. When tested alongside 26 radiologists, the model’s average performance exceeded that of 23 of the participating professionals. Furthermore, researchers noted that using the AI as a diagnostic aid resulted in a 10% reduction in missed diagnoses and a more than 30% reduction in diagnosis time.

Described by its developers as ‘the world’s first expert-level generalist medical imaging model,’ Damo Radar is intended to serve as a clinical aid rather than a replacement for radiologists. Alibaba has released the RADAR code and pretrained model resources publicly to facilitate further research.