TARGET aims to revolutionise the management of Atrial Fibrillation (AF) and AF-related strokes (AFRS). Despite extensive research and advancements in stroke prevention for AF patients, understanding the complex link between AF and stroke, as well as managing long-term risks, remains a challenge, posing substantial long-term risks such as stroke recurrence and bleeding complications.
By developing novel virtual twin-based AI models, TARGET combines mechanistic and data-driven virtual twins with causal AI to bridge the gap between research and clinical practice. These models consider established risk factors, comorbidities, imaging, and biomarkers to create personalized approaches that optimize stroke management, rehabilitation treatments, and enhance patients' quality of life. The integration of these models into monitoring devices and rehabilitation tools accelerates clinical adoption, reducing healthcare costs and overcoming challenges faced by healthcare systems.
Virtual twins in healthcare are digital representations of individuals created through data, analytics, and simulation. These personalised models provide a comprehensive view of a patient's health by integrating data from multiple organs and organ systems. They can improve clinical decision-making, optimize patient care, and enable predictive medicine by facilitating advanced measurements and in-silico therapy planning. TARGET will develop virtual twin-driven AI models and help consolidate existing mechanistic virtual twin models of the heart, the brain and the neuromusculoskeletal system, enriching these twins to deliver more complex tasks.
In TARGET patients and stakeholders will be involved in all stages of the clinical decision support tools design, ensuring patients and key stakeholders remain at the heart of the research and development of the project. TARGET will determine their understanding, acceptability and engagement with virtual twin models in healthcare. It will co-produce the clinical decision support tools and determine the potential barriers and facilitators of the implementation of these into clinical practice to promote end-user adoption.
Atrial fibrillation (AF) is an irregular and often very rapid heart rhythm (arrhythmia) that can lead to blood clots in the heart, it is the most common type of cardiac arrhythmia worldwide. AF is the most frequent cause of cardioembolic stroke, with increasing prevalence and incidence, especially in older patients. AF-relate stroke (AFRS) accounts for 20% of ischaemic strokes. AFRS is characterized by severe neurological deficits caused by larger blood clots, resulting in significant damage to substantial areas of the brain and potential bleeding. Furthermore, functional recovery from AFRS is frequently unsatisfactory.
Conventional assessment tools mostly focus on static risk factors and do not account for the dynamic nature of risk, which changes during acute illness, with ageing and depending on comorbidities. The modelling of virtual twins will allow for a dynamic risk assessment approach considering the rapid changes in biomarkers and vital signs to help optimise tools and reliably predict outcomes.
TARGET will incorporate biomarkers associated with cardiac or neurologic damage, along with automated analysis of ECG characteristics, to provide a more comprehensive understanding of the underlying pathophysiology. The integration of biomarkers and their dynamic changes into treatment models as well may support the formulation of individualised recovery plans to help reduce disability and enhance quality of life.
Novel technologies and digital solutions, such as the portable patch-based monitoring system (Isansys Lifetouch) with specific in-built AI, and the dynamic, personalised risk assessment tools based on virtual twins and AI will allow rapid incorporation of novel AI models into existing monitoring devices, such as bedside monitoring or medical wearable devices, as part of a new generation of patient monitoring systems. The involvement of industrial partners of the calibre of Siemens and Isansys will ensure that TARGET’s novel technologies and digital solutions are swiftly adopted by the healthcare sector, enhancing patient care, and accelerating clinical impact.
Focusing on patients at risk of atrial fibrillation (AF) and AF-relates stroke (AFRS), TARGET specifically targets three crucial areas for investigation and improvement:
of AF and major complications such as AFRS, including dynamic, longitudinal monitoring of AF and the subsequent risk of developing AFRS
of AFRS through TARGET’s virtual twin-based AI models that will help improve diagnostic and treatment processes for AFRS
focusing on identifying predictors of functional independence and quality of life in AFRS survivors and facilitating personalisation of rehabilitation
The coordinator and management team will oversee the execution of planned activities, the timely submission of deliverables and reports, and the financial and administrative management. WP1 will ensure the effective collaboration among partners, risk mitigation, and successful project implementation.
WP2 aims to evaluate stakeholder views, understanding, acceptability and engagement with virtual twin models in healthcare, supporting co-production to maximise usability, applicability, implementation, and dissemination of the virtual twin models and clinical decision support tools. It will also study tool utilization from both patient and HCP viewpoints and craft effective communication strategies to encourage adoption.
WP3 will provide the necessary tools and technology for dataset integration and sharing, with utilization and storage protected under current data-sharing frameworks. The objectives are focused on ensuring data security, enabling the integrated analysis of multi-source data for WPs 4-7, and supporting the application of new tools and models on clinical and synthetic data in WP8. Further, WP3 will validate the robustness and data quality of AI models to ensure their trustworthiness.
WP4 will advance translational research by building comprehensive multi-scale, multi-organ virtual twins that will form the foundation for AI modelling activities. The aim is to create mechanistic virtual twins of the heart, brain, and neuromuscular system, alongside data-driven counterparts using sophisticated, causative AI/ML techniques. These virtual twins will be dynamically upgradeable with new data and will integrate with the EDITH ecosystem to enhance collaborative research capacities.
WP5 is set to create virtual twin-based AI models tailored for individualized risk prediction of AF and AFRS. The WP will formulate detailed risk prediction models within multi-dimensional virtual twins, pinpointing potential biomarkers for these conditions. The virtual twin-based AI models will be integrated into Isansys’ Patient Status Engine and will be readily available in multiple countries. Furthermore, it will deliver decision support tools designed for updating the risk predictions dynamically and for the ongoing monitoring of AF.
WP6 focuses on mapping the journey of patients after experiencing a stroke. It aims to build predictive models for determining stroke causes, risks of HT or recurrence at an early stage, and short to long-term outcomes post-stroke. Strategies include using virtual twins for stroke aetiology identification, creating AI/ML models for predicting patient recovery trajectories, and developing a support tool for acute stroke diagnosis and treatment planning.
WP7 aims to develop advanced AI models based on virtual twins that will personalize care for AFRS survivors by identifying predictors of functional independence and quality of life. These models will enable predictions of patient outcomes following rehabilitation and recognize individual needs to tailor care. Moreover, WP7 seeks to refine rehabilitation programs through innovative tools for continuous assessment and interactive applications.
WP8 Clinical studies and in-silico trials led by Liverpool University Hospitals NHS Foundation Trust
WP8 will perform prospective cohort studies at four EU centres. The data collected will be instrumental in validating the performance of virtual twin-based AI models. Additionally, this data will facilitate in-silico trial simulations aimed at assessing the efficacy of integrating such models into clinical practice compared to the standard of care. Results will provide a foundation for evidence-based guidelines and support broader implementation.
WP9 aims to effectively communicate project ideas, activities, and results to diverse audiences, influencing healthcare policy, patient outcomes, and service delivery. This entails spreading information about the activities and outcomes to relevant stakeholders, including patient groups, healthcare professionals (HCPs), and the scientific community, in addition to devising strategies for leveraging results through plans aimed at gaining regulatory approval. Work Package 9 (WP9) will investigate alternative strategies for applying scientific breakthroughs to achieve a lasting influence. It will also participate in collaborative efforts with other initiatives to promote the exchange of knowledge and the grouping of activities. The collaboration with Isansys and SHS will play a crucial role in utilizing the findings of TARGET and will provide essential guidance to academic and clinical partners. This ensures that their development strategies across various work packages comply with the Medical Device Regulation (MDR).