From the Journals

AI Shows High Accuracy in CT, MRI Protocoling

Meta-analysis finds similar protocoling performance across machine learning, BERT-based models, and large language models.

Share

A systematic review and meta-analysis of 23 studies involving about 1.2 million imaging orders revealed that artificial intelligence systems accurately assign CT and MRI protocols with an overall accuracy of 85%. Models such as traditional machine learning, transformer-based technologies, and large language models were evaluated, showing accuracy rates between 65% and 95%. Investigators find potential for AI to enhance radiology workflows but note that ambiguous requisition text and data imbalance can lead to errors. Future research should explore clinical trials and refined AI models.

Original Source(s)

AI Shows High Accuracy in CT, MRI Protocoling

  • by Doug Brunk

  • March 9, 2026

Related Content

Neurointerventions Face Waste Audit

Investigators quantify procedural waste, disposal costs, and carbon emissions in a neurointerventional angiography suite.

Guideline Gap in Facial Trauma Imaging

Retrospective study highlights practice variation and continued use of low-value radiographic imaging for maxillofacial trauma evaluation

Can MRI Predict Perforator Stroke Progression?

Investigators evaluate multisequence imaging for its ability to identify patients at risk for neurologic progression.