AI-POWERED HEMODIALYSIS MONITORING & RISK PREDICTION
Integrating hemodialysis device data, hospital information systems, and AI-powered analytics to support healthcare professionals in monitoring dialysis treatment, assessing patient conditions, and identifying potential risks for clinical decision support.
Joint Exhibition｜MuSheng Technology Co., Ltd. × Taipei Veterans General Hospital
Technology Development & Clinical Collaboration｜Taipei Veterans General Hospital
Product Development & Commercialization｜MuSheng Technology Co., Ltd.
.png)
Dialysis Records → Data Integration → AI Analysis → Abnormal Alerts → AI Recommendations
THREE CORE VALUES
Real-time Monitoring
Monitor dialysis parameters and patient status throughout treatment.
Data Integration
Integrate dialysis machines, HIS, LIS, PACS, and standardized FHIR data across systems.
AI Risk Prediction
Analyze dialysis data and provide AI-assisted heart failure risk alerts for clinical attention.
Four AI Clinical Prediction Models
REAL-TIME MULTI-PATIENT MONITORING
Monitor dialysis status across multiple patients and identify potential risk levels at a glance.
Color-coded risk levels (Normal / HF / IDH) help clinicians spot cases needing attention.
Clinical Validation & Real-World Evidence
Two-Year Clinical Evidence at Taipei Veterans General Hospital
Observed outcomes before and after deployment of the heart failure risk prediction model.
5.94% → 3.93%
Mortality rate
14.87% → 7.45%
One-year mortality
5.40% → 4.23%
CHF-related mortality
79.31% → 45%
Cardiopulmonary-system mortality
Multi-Level Clinical Validation — Heart Failure Risk Prediction Model
Validated with clinical cases across a medical center, regional hospitals, and dialysis clinics. Model performance is shown below.
Sensitivity 0.833｜Specificity 0.947｜AUC 0.809
