Jump to page content

Quality AssuranceOngoing

STRATIF-AI

Continuous stratification for improved prevention, treatment, and rehabilitation of stroke patients using digital twins and artificial intelligence (AI).

A platform that integrates mechanistic modelling and bioinformatics to create a real-time virtual copy of a patient, allowing for detailed health simulations. The project connects multiple apps to track a stroke patient’s individual journey, covering all phases, from prevention to treatment and rehabilitation.

Background / Rationale

Traditional patient stratification relies often on periodic assessments, which misses the rich, continuous data generated between appointments. Data, sourced from diverse technologies, medical specialties, and disease stages, must be integrated to create a complete patient profile and ultimately improve healthcare decisions. The STRATIF-AI project introduces a novel approach—continuous stratification—which utilizes all available patient data in a Personal Data Vault to create a continuously updated digital twin. This twin incorporates both traditional data sources, e.g., MRI images, blood samples, and new sources like wearable devices and patient diaries. By combining mechanistic models with machine learning (ML) and bioinformatics, this approach ensures a comprehensive, real-time view of each patient’s health, aiming to improve care across all phases of stroke management: prevention, acute treatment, and rehabilitation.

Objectives

This project includes following objectives:

  • Data Harmonization and Integration: Develop a system to semantically harmonize and integrate diverse stroke-relevant data sources.

  • Model Development and Validation: Enhance digital twin models to simulate patient-specific physiological processes.

  • STRATIF-AI Platform Development: Co-design a stroke stratification platform, integrating existing apps and input from clinical end-users.

  • Clinical Impact Assessment: Conduct Effective Implementation Hybrid trials comparing patient outcomes with and without the technology.

  • Trustworthy AI and Data Privacy: Ensure Trustworthy AI principles are incorporated, aligning with EU guidelines and regulations.

  • Dissemination and Generalization: Organize activities that disseminate and exploit the project’s results to a wide range of stakeholders.

Methodology

The project will employ a Federated Learning platform to manage data from multiple sources under stringent legal and security conditions. A data harmonization and interoperability system will ensure that diverse data types are fully integrated and accessible within a Personal Data Vault. This hybrid digital twin platform will serve as a backend to all frontend applications, which will be co-designed with stakeholders to ensure ethics and trustworthiness across stroke prevention, treatment, and rehabilitation phases.

Expected Results / Implications / Perspectives

The STRATIF-AI project aims to transform healthcare by introducing continuous stratification, a process which will help to bridge gaps in traditional intermittent assessments. Scientifically, the platform boasts state-of-the-art data integration and analysis methods, enhancing clinical research and patient stratification. Economically and technically, it offers expandable capabilities, attracting new actors like wearable sensor companies and facilitating the adoption of evidence-based guidelines. Societally, STRATIF-AI represents a shift towards a patient-centric healthcare ecosystem, promoting continuous stratification and personalized care. Its potential impact spans improved patient outcomes, reduced disease burden, and enhanced trust in innovative healthcare technologies.

Further Information

Funders and Cooperation Partners

This project has received funding from the European Union’s Horizon 2022 programme under Grant Agreement No. 101080875.
 

Project Partners

Contact