Health virtual twins for the personalised management of stroke related to atrial fibrillation

Introducing virtual twin-driven AI models for atrial fibrillation (AF) and its complications such as AF-related stroke (AFRS). Novel personalised virtual twins and decision-support tools to help prevent AF and AFRS, while optimising acute management and rehabilitation. Ultimately, the goal is to provide a better quality of life for patients and caregivers, and lower healthcare costs.

With AF being a widespread irregular and often very rapid heart rhythm (arrhythmia), it significantly increases the risk of complications such as stroke and heart failure. TARGET represents a significant milestone by reshaping risk prediction, diagnosis, and management of AF and AFRS, and accelerating the translation of research into practical applications.

Integrated, multi-scale virtual twins modelling

Artificial Intelligence (AI) solutions are being increasingly used in clinical practice. Coupled with AI, virtual twins aim to improve patient outcomes by enabling more accurate diagnoses, more personalised treatment plans, and more efficient resource allocation.

PATIENT-CENTERED

TARGET has a strong focus on the personalisation of health technologies for improved and more cost-efficient solutions in disease prevention, diagnosis, treatment and rehabilitation.

DIGITAL

Virtual twins are digital representation of individuals based on their specific characteristics, and medical and health status history.

DYNAMIC

TARGET accounts for the dynamic nature of risk and includes biomarkers, moving beyond models that focus mainly on static risk factors.

The workplan

TARGET represents a milestone project to improve the care of patients in the AFRS pathway, introducing a paradigm shift in risk prediction, diagnosis and management of the disease. The project’s workplan is structured around:

Novel virtual twin-based AI models and mechanistic virtual twins

TARGET will develop integrated, multi-scale computational models (virtual twins) and tools for personalised risk prediction, optimised management and treatment of stroke, and enhanced rehabilitation. TARGET will also help consolidate existing mechanistic virtual twin models of the heart, the brain and the neuromusculoskeletal system, enriching these twins to deliver more complex tasks, and supporting research to move towards a more integrated human virtual twin.

Co-creation of the clinical decision support tool

To ensure stakeholders’ involvement in the design and development of the decision support tools a co-creation group will be established. The co-production includes four phases:

  1. open-design: patients, members of the public, HCPs, and healthcare commissioners will be included to gather their views, understanding, acceptability, and engagement with virtual twin models for healthcare
  2. co-design: participants will explore current challenges of using virtual twins for clinical decision-making, utilising information from the discussion groups and exploring the potential solutions for clinical decision support tools
  3. co-development: together with software designers will develop the five clinical decision support tools, which are fit-for-purpose and have a health equity focus
  4. co-customisation: participants will iteratively test prototypes of the clinical decision support tool, employing a shared prioritisation process to select the ‘best’ ideas

Validation of models and tools

Before wider implementation into clinical practice, external validation of the novel models and tools is required. The project consortium will collect new observational data via 4 carefully designed prospective clinical studies, which will be used to test and validate the personalised tools and the virtual twin models using a clinical trial simulation (virtual/in-silico), to demonstrate evidence of clinically meaningful results.