Deep learning AI improves interreader agreement in CAD-RADS assessments on coronary CT angiography.
Key Details
- 1Study published in Radiology: Cardiothoracic Imaging (Oct 9).
- 2623 patients (11,214 coronary segments) who underwent CCTA were included.
- 3AI-assisted CAD-RADS readings showed higher agreement: 1 vs. 2 readers (manual: 73.7%, AI-assisted: 81.8%); 1 vs. 3 (manual: 77.9%, AI-assisted: 84.5%).
- 4AI algorithm evaluated lesion, segment, and patient levels for CAD-RADS and classified stenosis grades with visual outputs.
- 5Authors call for further research with larger datasets and diverse settings.
Why It Matters
Standardizing CAD-RADS assessments using AI has the potential to reduce variability and errors in coronary CT reporting, which could improve patient management and facilitate training for less experienced readers.
Source
AuntMinnie
Related News
•HealthExec
AI's Expanding Role in Healthcare and Implications for Radiology
A series of thought leaders and institutions weigh in on AI's transformative potential in healthcare, with emphasis on radiology adoption and responsible use.
•Radiology Business
Most Radiology AI Users Lack Clear Evidence of Financial ROI
Survey finds over 75% of radiology organizations using AI lack clear, quantified ROI data.
•AuntMinnie
Toronto Study: LLMs Must Cite Sources for Radiology Decision Support
University of Toronto researchers found that large language models (LLMs) such as DeepSeek V3 and GPT-4o offer promising support for radiology decision-making in pancreatic cancer when their recommendations cite guideline sources.