This session focuses on the integration of authoritative evidence-based medical knowledge bases, rule engines, natural language processing technologies, real-world clinical data, best clinical practices, clinical predictive models, and closely integrated clinical workflows to develop the core technology solution for intelligent CDSS. The approach aims to streamline the clinical decision-making process for clinical diagnosis and real-time care, improve patient clinical transition, and make the clinician's treatment process more efficient, standardized, safe and reliable.
* Clinical Data Modelling: research on integration methods, standardization methods and data-driven machine learning methods for various types of patient data, and research on predictive models for clinical applications.
* Knowledge Modelling: research on integration and modelling methods of clinical knowledge such as clinical pathways, clinical guidelines, medical literature, best clinical practices, etc., research on logical reasoning methods, research on methods of expert knowledge base construction.
* Human-computer Interaction Models: research on personalized interaction models for unused user groups such as physicians, nursing staff, laboratory staff and patients, and research on clinical workflow integration solutions for CDSS.
The correct information is provided to the appropriate person, at the proper time and in the exact right mode of intervention, through the rational channels in the consultation process.
Clinical Prediction and Evaluation in the ICU Scenario:
-- Clinical Phenotyping of Diseases
-- Early warning model of AKI
-- Early warning model of sepsis
-- Mechanical ventilation related predictive model
-- Drug Dose Prediction Model for Critically Ill Patients
-- Models assessing disease severity and mortality prognosis
-- Other predictive models for clinical management
-- Knowledge Base Construction Based on Clinical Guidelines