New EHR framework improves perioperative outcome prediction

Researchers have developed a new domain-structured ensemble framework designed to predict perioperative outcomes using electronic health record (EHR) data. This framework organizes predictors into patient, surgery, and anesthetics domains, with domain-specific gradient boosting models integrating their risk estimates via a logistic regression meta-learner. Tested for predicting postoperative delirium, the system achieved an AUROC of 0.899, outperforming single-stage models and demonstrating excellent calibration and temporal validation. AI

IMPACT This framework offers a scalable foundation for interpretable, calibration-aware perioperative clinical decision support systems.

RANK_REASON The item is a research paper detailing a new framework for outcome prediction using EHR data. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv
• Confusion Assessment Method
• dementia
• electronic health records
• gradient boosting
• logistic regression model
• Postoperative delirium

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →