The SIATE project uses Artificial Intelligence (AI) to monitor and personalize hemodialysis therapy in real time: developing a digital platform that captures clinical data, analyzes it with AI, and helps clinicians tailor therapy to each patient, integrating smart monitoring, mobile and telemedicine tools, and decision support systems to enable personalized, preventive kidney care.
Patients with chronic kidney disease undergoing dialysis often face serious health risks due to metabolic imbalances and inefficient treatment
Move from reactive treatment to predictive, AI-guided care that reduces complications, improves vascular access management, and enhances quality of life
AIM
- To develop an AI-powered system that enhances the quality, precision, and accessibility of hemodialysis therapy by:
- Automatically monitoring vascular access (VA) performance through early alert generation, enabling preventive rather than corrective interventions
- Supporting clinical decisions by assessing metabolic and nutritional control on a per-patient basis
- Estimating dialysis efficiency via innovative sensors capable of simultaneously measuring standard and novel biomarkers in blood, dialysate, and exhaled air
- Standardizing therapies and objectively recording metabolic events for consistent clinical intervention
- Expanding the use of home dialysis through telemonitoring and AI-guided therapy adjustments
- Creating large-scale, structured databases for Network Medicine and AI applications in both clinical and market-oriented research
Method
Nine implementation phases (Objectives – OR):
1) Upgrade and automation of the existing VA triage system, initially manual, to serve 1,500 hemodialysis patients with real-time data acquisition and adaptive learning capabilities
2) Design of a clinical data acquisition and storage system to feed AI algorithms for decision-making model training
3) Development of a Clinical Decision Support System (CDSS) that integrates AI-driven decision rules into clinical workflows
4) Creation of a predictive system, based on clinical data, for personalized risk assessment and therapeutic planning
5) Implementation of a smart infrastructure to collect data from dialysis monitors and common devices (e.g., bioimpedance analyzers, blood gas machines, scales), enabling AI-based forecasting of clinical events
6) Deployment of a cloud-based infrastructure for storing and processing clinical data, supporting CDSS operations and AI model evolution
7) Design and release of a Data Portal for Medical Devices & Patient Data, making clinical and monitoring data available to AI processing modules
8) Analytical method development for measuring both known (targeted) and novel (untargeted) biomarkers in plasma, dialysate, and breath samples
9) User-centered design and ecosystem engagement, including application design feedback, stakeholder education, and simulation modeling with representative patient cohorts
Results
- Generate substantial clinical and operational improvements
- Automated, multi-parametric, and personalized monitoring of dialysis patients
- Scalable solution for global implementation in all hemodialysis patients
- Enablement of remote monitoring and home-based dialysis through telemedicine
- Improved vascular access management, fewer complications and hospitalizations, better patient outcomes
- Enhanced quality of life and reduced healthcare costs through earlier interventions and more accurate treatment delivery
- Introduction of new clinical parameters and expanded monitoring capabilities
Conclusion
SIATE represents a paradigm shift in dialysis treatment, placing AI at the heart of clinical decision-making
The system’s capacity to integrate real-time data from a wide range of sources, apply predictive analytics, and support individualized treatment planning makes it a powerful tool for improving outcomes in a fragile patient population; its focus on scalability, automation, and telehealth ensures broad applicability across healthcare settings, supporting the move toward precision nephrology and value-based care