A new radiology model tackles 146 abdominal scan findings
A study published in Science on September 17 introduces RADAR, an AI model for interpreting abdominal CT scans. The researchers report a mean AUC of 0.913 across 146 findings. The best competing vision-language model in their evaluation scored 0.776.
Verified 12:21 AM PDT · 2 original sources
AUC measures how well a model separates positive and negative cases. It is not the percentage of patients it diagnoses correctly. The study reports an AUC of 0.895 across cohorts from eight outside centers.
The results come from the researchers' evaluation. They do not show that routine use improves patient outcomes or that RADAR can replace a radiologist. Others can inspect the published code, but that does not amount to independent clinical validation.
Independent evaluations and studies in working hospitals could show where RADAR helps. Those studies need to count missed findings and false alarms. They also need to test whether clinicians make better decisions.
Audit the story
Original sources
Company claims remain company claims. Follow the reporting and judge the evidence directly.
- American Association for the Advancement of ScienceIntroducing RADAR, a generalist AI tool for abdominal CT diagnosis ↗
- Alibaba DAMO AcademyRADAR: An expert-level generalist AI for abdominal CT diagnosis ↗
Continue the edition