CVDLINK

A federated paradigm of real-world data sources utilization for the empowerment of diagnosis, prognosis and risk assessment of cardiovascular conditions

The CVDLINK project is dedicated to addressing the significant challenge of cardiovascular disease (CVD) within the EU, where over 6 million new cases and 1.8 million deaths are reported annually. Recognizing the urgent need for effective prevention strategies, CVDLINK employs a scientific approach aimed at transforming the management, treatment and research of cardiovascular disease through data-driven innovation. The project aims to tackle challenges related to data availability and management by implementing a privacy-by-design European-wide federated platform-as-a-service (PaaS). The goal is to deliver effective data-driven human-centric interventions and the advancement of research and management in the CVD domain. An essential part of the project is the development and implementation of AI and data-powered precision medicine tools and pipelines. These tools are designed to enhance diagnosis, risk stratification, and treatment protocols for seven distinct cardiovascular conditions, leveraging diverse data from retrospective datasets, cohort studies, and biobanks across seven countries. This endeavor aims to set a new standard for utilizing heterogeneous data in crafting advanced AI-driven tools, thereby offering significant benefits to health systems, patients, the industry, and EU citizens at large. The efficacy of these tools will be properly validated in five countries to demonstrate CVDLINK’s impact. Furthermore, the project will produce a set of best practices, conduct a cost-effectiveness analysis, and engage in systematic awareness campaigns to encourage widespread adoption.

M4D will contribute to the research and development of XAI techniques aimed at enhancing the trustworthiness of the developed AI algorithms. Specifically, M4D will enhance AI algorithms with the ability to provide understandable explanations for users. To achieve this, different XAI post-hoc techniques will be applied according to the type of AI models implemented i.e model-specific or model-agnostic techniques. Furthermore, M4D will lead the development of appropriate visualizations to provide a set of interfaces for the developed algorithms and outcomes that will allow the use of the CVDLINK tools, providing user services that combine AI models with existing scientific knowledge. Finally, M4D serves as the Quality Manager of the project.