Researchers validate the REDMOD AI framework, which detects early, otherwise invisible, tissue changes of pancreatic cancer on CT scans before clinical diagnosis.
Key Details
- 1REDMOD is an AI radiomics model for detecting subtle texture changes indicative of stage 0 pancreatic ductal adenocarcinoma on CT.
- 2The model identified preclinical disease an average of 475 days before clinical diagnosis.
- 3In head-to-head comparison, REDMOD's sensitivity was nearly double that of expert radiologists (73% vs. 39%).
- 4Performance remained high (68% vs. 23%) for cancers detected over two years before diagnosis.
- 5The model showed strong generalisability in external cohorts (correctly identifying >81% and 87.5% of cases in independent datasets).
- 6Researchers caution further prospective validation and testing in high-risk patients are needed before clinical adoption.
Why It Matters
Ultra-early detection of pancreatic cancer could dramatically improve survival rates by enabling treatment before symptoms arise. AI models like REDMOD may soon provide radiologists with powerful new tools to catch deadly cancers in their most treatable phases.
Source
EurekAlert
Related News
•EurekAlert
AI-Enhanced Imaging Breaks Barriers in Complex Media for Biomedical Use
Researchers developed a physics-based machine learning imaging system that enhances visibility through complex media, holding promise for biomedical imaging.
•EurekAlert
AI Model Accelerates Accurate Dose Prediction for CyberKnife Radiotherapy
A physics- and spatially informed diffusion model accelerates Monte Carlo–level dose estimation for CyberKnife radiotherapy using planning CT and anatomical data.
•EurekAlert
New G-AUDIT Tool Detects Hidden Bias in Medical AI Training Data
Johns Hopkins and FDA researchers unveil G-AUDIT, a tool to identify hidden biases in medical AI datasets before model training.