Prediction models using AI and statistical methods are being refined across healthcare domains, from lung and breast cancer screening to liver cancer risk assessment. New research reveals significant accuracy gaps across racial and ethnic groups, while fresh AI frameworks enable models to compete and improve on real-world data.
·Lung cancer prediction models show varying accuracy depending on patient race and ethnicity, raising equity concerns in screening.
·AI-powered breast cancer risk assessment now combines mammographic imaging with predictive modeling for stronger detection.
·New zero-shot foundation models for tabular data and AI connectors are making prediction tools more accessible across platforms.
·Risk-based screening strategies outperform current US preventive guidelines for identifying high-risk lung cancer candidates.
·Competitive AI frameworks now allow prediction models to earn rewards and improve based on real-world forecasting performance.
drawn from IARC – INTERNATIONAL AGENCY FOR RESEARCH ON CANCER, PR Newswire, respiratory-therapy.com, Inside Precision Medicine · updated 9h ago