Explainable ai system for ICU diagnosis prediction

An explainable AI system using ensemble and deep learning models with ClinicalBERT enhances ICU diagnosis prediction by analyzing electronic health records, improving accuracy and transparency for timely clinical decisions.

WO2026143167A1 Publication Date: 2026-07-02NORTHEASTERN UNIV (US)
5 Cites 0 Cited by

Patent Information

Application Number
PCT/US2025/061223
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2024-12-26
Filing Date
2025-12-23
Publication Date
2026-07-02

AI Technical Summary

Technical Problem

Existing clinical decision support systems fail to accurately predict missed severe diagnoses in intensive care units due to the lack of leveraging vast patient data and providing transparent, explainable outputs, limiting clinician trust and effective decision-making.

Method used

An explainable AI system utilizing ensemble models, deep learning models, and large language models to analyze electronic health records, generating interpretable explanations for predicting conditions like acute myocardial infarction, pulmonary embolism, and others, by combining logistic regression, random forest, XGBoost, and support vector machines, with ClinicalBERT for unstructured clinical notes, and providing real-time risk assessment.

Benefits of technology

Improves prediction accuracy and clinician trust by capturing diverse clinical patterns and providing transparent reasoning, enabling timely and informed decision-making in ICU environments.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

An explainable artificial intelligence system for early prediction of missed severe diagnoses in intensive care unit patients is provided. The system comprises a data processing module configured to receive and process electronic health record data including patient demographics, vital signs, laboratory results, and diagnostic codes. A machine learning module with ensemble, deep learning, and large language models analyzes processed data to predict likelihood of severe diagnoses including acute myocardial infarction, acute pulmonary embolism, bacterial pneumonia, aortic dissection, meningitis, spinal abscess, bacterial endocarditis, and arterial thromboembolism. An explainability module generates feature importance rankings and provides interpretable explanations for predictions.
Need to check novelty before this filing date? Find Prior Art