A new AI model outperformed radiologists in identifying difficult-to-detect femoral neck fractures on radiographs.
Key Details
- 1The AI tool, OccuNet, was trained and validated to detect femoral neck fractures, a challenging diagnostic area for radiologists.
- 2OccuNet uses a two-stage approach: contrastive pre-training on paired original and artifact-augmented radiographs, followed by fracture detection fine-tuning.
- 3The model was tested on a cohort of 2,576 adults with suspected hip trauma across four hospitals.
- 4Studies indicate that up to 10% of femoral neck fractures can be missed on initial radiographs, leading to delayed treatment and complications.
- 5The study was published in RSNA’s journal Radiology and involved experts from the Zhejiang Spine Research Center, China.
Why It Matters
Missed or indeterminate hip fractures can result in significant patient harm, so an AI model offering improved sensitivity and performance offers a valuable safety net. Such AI tools could enhance diagnostic accuracy, reduce missed injuries, and potentially improve patient outcomes in emergency radiology settings.
Source
Radiology Business
Related News
•Radiology Business
Healthcare Leader Warns AI Will Dominate Diagnostic Radiology
A leading oncologist urges future radiologists to specialize in interventional procedures due to AI advances in image interpretation.
•Radiology Business
Harrison.ai Restructures and Launches U.S. AI-Enabled Teleradiology Venture
Harrison.ai lays off staff amid major restructuring and launches Frontier Radiology, an AI-powered U.S. teleradiology business.
•Radiology Business
UCLA Review Evaluates Breast AI for Early Cancer Detection
A UCLA-led review analyzes current evidence for AI tools in screening mammography, focusing on their capability to detect interval cancers.