AI predicts brain tumor enhancement from non-contrast MRI, outperforming radiologists

Researchers have developed a deep learning model capable of predicting brain tumor enhancement from non-contrast MRI scans, potentially reducing the need for contrast agents. The model, trained on over 11,000 studies, achieved 83.0% balanced accuracy in detecting enhancement, outperforming eleven expert radiologists who achieved 71.7% accuracy under blinded conditions. This AI shows promise as a decision-support tool, particularly for flagging studies likely to show enhancement and lessening gadolinium dependence in neuro-oncology. AI

IMPACT This AI model could streamline neuro-oncology imaging workflows and reduce patient exposure to contrast agents.

RANK_REASON The cluster is based on a research paper detailing a new AI model and its performance on a diagnostic task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• artificial intelligence
• deep learning model
• gadolinium
• nnU-Net
• radiologists

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →