AI Model Detects Pancreatic Cancer Years Before Diagnosis
- Mayo Clinic’s REDMOD AI model achieved a detection rate of 73% for pancreatic cancer on routine CT scans, median of 16 months before clinical diagnosis.
- The AI analyzed nearly 2,000 CT scans initially interpreted as normal, including those from patients later diagnosed with pancreatic cancer.
- REDMOD’s detection capability nearly doubles that of radiologists, who had a 39% detection rate on the same scans.
- The AI detected pancreatic cancer an average of 475 days before clinical diagnosis with an accuracy of 88%.
- For cases identified more than two years prior to clinical diagnosis, REDMOD was nearly three times as accurate as human specialists, achieving a 68% accuracy rate compared to 23%.
Mayo Clinic’s REDMOD AI model represents a significant advancement in early detection of pancreatic cancer by identifying subtle changes in routine CT scans long before symptoms appear. This capability is crucial as pancreatic cancer is projected to become the second-leading cause of cancer-related death in the U.S. by the year specified in data projections.
REDMOD’s ability to detect pancreatic cancer up to three years before clinical diagnosis could significantly improve patient outcomes by allowing earlier intervention when treatment options are more effective (Source)
