Overview
Mayo Clinic researchers have developed REDMOD (Radiomics-based Early Detection MODel), an AI system capable of detecting pancreatic cancer from routine CT scans an average of 475 days before clinical diagnosis. The study was published in the journal Gut in April 2026.
Key Findings
- Detection Timeline: REDMOD identified pancreatic ductal adenocarcinoma (PDAC) an average of 475 days (approximately 15 months) before clinical diagnosis
- Accuracy: 73% accuracy in detecting early-stage changes, compared to 39% for experienced radiologists
- Long-term Performance: For scans taken more than 2 years before diagnosis, REDMOD maintained 68% accuracy while radiologist detection dropped to 23%
- Technology: Uses radiomics to analyze subtle tissue texture patterns imperceptible to the human eye
Clinical Significance
Pancreatic cancer is often diagnosed at advanced stages due to difficulty of early detection, resulting in a 10-13% 5-year survival rate. Researchers suggest REDMOD could significantly improve survival rates by enabling earlier intervention, potentially increasing localized detection from 10% to 50%.
Research Team
The study was led by Sovanlal Mukherjee, PhD, and colleagues at Mayo Clinic. Prospective clinical trials are required before implementation in routine medical practice.
Sources
- auntminnie.com
- straitstimes.com
- dataconomy.com
- forklog.com
- letsdatascience.com
- bioengineer.org
- news-medical.net