A new study published in Nature reveals that AI systems can now detect early signs of pancreatic cancer from routine CT scans with significantly higher accuracy than human radiologists. The research, led by a team from MIT and Harvard Medical School, analyzed thousands of scans and identified subtle patterns invisible to the human eye. The model achieved a 94% detection rate for stage 1 and 2 cancers, compared to the current average of 69% for radiologists. This advancement could lead to earlier interventions and improved survival rates for a disease often diagnosed too late. The researchers emphasize the tool is designed to assist, not replace, clinicians, and further trials are needed before clinical deployment. Read the full article at https://technologyreview.com/2024/05/15/ai-pancreatic-cancer-detection-study.