Related Experiment Video
Updated: May 16, 2026

A Data Integration Workflow to Identify Drug Combinations Targeting Synthetic Lethal Interactions
Published on: May 27, 2021
AI-WAR: a novel warfarin management software with a bidirectional LSTM dosing model improves time in therapeutic
1Center of Clinical Pharmacology, The Third Xiangya Hospital, Central South University, Changsha, China.
An artificial intelligence system (AI-WAR) improved warfarin management by increasing time in therapeutic range and reducing adverse events. This AI tool enhances anticoagulation quality and supports remote patient monitoring in primary care.
Area of Science:
- Cardiovascular Medicine
- Artificial Intelligence in Healthcare
- Pharmacogenomics and Precision Medicine
Background:
- Warfarin is a critical anticoagulant for heart valve replacement patients but poses management challenges due to its narrow therapeutic window and monitoring needs.
- Suboptimal anticoagulation quality is common in real-world settings, especially in resource-limited primary care.
- Interindividual variability and frequent INR monitoring complicate effective warfarin management.
Purpose of the Study:
- To develop and validate an artificial intelligence-based warfarin management system (AI-WAR) for improved anticoagulation.
- To assess the efficacy of AI-WAR in enhancing anticoagulation quality and reducing adverse events compared to conventional methods.
- To evaluate the predictive performance of LSTM and Bi-LSTM models for individualized warfarin dosing.
Main Methods:
- Developed an AI-WAR system integrating remote follow-up, data management, and individualized dosing prediction models.
- Trained models using data from a randomized controlled trial (n=624) and a prospective registry (n=176).
- Validated the system on an independent real-world cohort (n=200) and compared outcomes with conventional management.
Main Results:
- AI-WAR significantly improved median time in therapeutic range (TTR) from 48.7% to 81.3% (P < 0.001).
- AI-WAR increased time in target INR (36.3% vs. 61.9%, P < 0.001) and reduced adverse events (9.6% vs. 19.7%, P < 0.001).
- The Bi-LSTM model demonstrated superior dose prediction accuracy (80.3%) and stable dose prediction (93.2%), reducing overdose predictions.
Conclusions:
- AI-WAR, coupled with Bi-LSTM models, serves as an effective decision-support tool for precision warfarin dosing.
- The system enhances anticoagulation quality, minimizes adverse events, and offers practical benefits for remote and primary care settings.
- AI-WAR supports the broader implementation of precision anticoagulation therapy, improving patient outcomes.
Related Concept Videos
Anticoagulant Drugs: Vitamin K Antagonists and Direct Oral Anticoagulants
Warfarin, a prominent vitamin K antagonist family member, exerts its effect by inhibiting the enzyme VKORC1 (vitamin K epoxide reductase complex 1). By hindering this enzyme, warfarin...
Therapeutic Drug Monitoring: Affecting Factors
Anticoagulant Drugs: Low-Molecular-Weight Heparins
Venous Thrombosis III: Interprofessional Care
Dosage Regimens: Designs and Approaches
Dosage Regimens: Partial Pharmacokinetic Parameters
