Stanford researchers introduce 'Merlin,' a versatile AI model that excels at various diagnostic tasks on 3D abdominal CT scans.
Key Details
- 1Merlin is a vision-language model trained on over 15,000 abdominal CT scans, more than 6 million images, and over 1.8 million diagnosis codes.
- 2It is designed for multiple diagnostic tasks, including anatomical abnormality detection and long-term disease prediction.
- 3The model's foundational approach allows for further expansion of its capabilities with minimal fine-tuning.
- 4Researchers highlight that Merlin could streamline radiology workflows and aid in clinical decision-making.
- 5The National Institutes of Health and NIBIB supported this project, which reportedly utilizes the largest abdominal CT dataset to date.
Why It Matters
Merlin demonstrates the potential of large, multi-task AI models to improve efficiency and accuracy in radiology, moving closer to comprehensive opportunistic diagnostics from CT imaging. Its success validates the push towards foundational models in medical imaging, promising scalable and flexible AI tools for clinical use.
Source
Radiology Business
Related News
•AuntMinnie
Radiology Receives Declining Share of Industry Research Funding
Radiologists received only 1.1% of industry-funded research payments in 2024, with a continuing downward trend.
•AuntMinnie
GPT-4o AI Matches Radiologists in Follow-Up Imaging Recommendations
GPT-4o matched the performance of experienced radiologists and surpassed residents in recommending follow-up imaging from routine radiology reports.
•Cardiovascular Business
AI Leverages Head CTs for Automated Heart Risk Assessments
AI models can turn routine head CT scans into automated cardiovascular risk assessments, expanding the utility of radiology studies.