A self-supervised AI model predicts 10-year mortality and hip fractures using only DEXA scans.
Key Details
- 1AI model uses self-supervised vision transformer architecture on DEXA images.
- 2Trained on 85,461 DEXA scans with no clinical or demographic data inputs.
- 3Tested on 17,218 whole-body scans (6,911 mortalities) and 4,199 hip scans (258 fractures).
- 4Achieved AUROC of 0.7 for mortality prediction, 0.74 for hip fracture prediction.
- 5Model requires no additional imaging or patient burden, utilizing routine DEXA exams.
Why It Matters
This advance highlights how deep learning can extract prognostic information from standard imaging, potentially transforming routine DEXA exams into powerful predictive tools that enhance patient care without added burden.
Source
AuntMinnie
Related News
•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
Paul Chang Discusses Foundation Models and Agentic AI at RSNA 2025
Dr. Paul Chang shares his insights on the role of foundation models and agentic AI in radiology at RSNA 2025.
•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.