Sign in
Ctrl K


Improving the vantage6 federated learning ecosystem

The amount of information generated in the domain of health care, like in many others, opens a wide range of research and development opportunities, such as the ones supported by prescriptive, predictive, and descriptive analytics. However, data privacy rules and regulations in the health domain tend to be stricter, as data breaches or unauthorized disclosure of data (e.g., health records, diagnoses, treatments) have significantly serious consequences for the individuals involved. Consequently, it has been challenging to fully exploit data in this domain as most of it sits in data silos with highly constrained access policies. Given these constraints, the Federated Learning (FL) paradigm has emerged as a promising approach for generating machine learning (ML) models, without disclosing patient data, while keeping the data at its original location.

FL is still a relatively new paradigm in machine learning, with technical barriers that are limiting broader adoption by researchers, not only in different medical specialties but also domains beyond the health-related ones. This proposal outlines a plan to allocate the joint generalization budget (granted to projects on the early detection and prevention of cardiovascular diseases, under the Big Data & Health call) to activities aimed at reducing these barriers. In particular, we identified Vantage6, an open-source software with an active community for the implementation of FL architectures, as the most promising tool for the facilitation of FL adoption. Consequently, this plan involves making a significant contribution to the evolution of vantage6 and, thus, the uptake of FL approaches. The motivation for this choice is twofold. On the one hand, vantage6 already provides key features for implementing FL in real settings. It offers FL-related autonomy and privacy features that enable each party to control their own data. Additionally, it provides flexibility in terms of programming languages and data partitioning models and enables the integration of heterogeneous nodes in terms of the hardware and operating systems. On the other hand, the most important advantage of vantage6 is its growing community of developers and researchers, who have adopted it in health-related projects (including eScience Center ones) and have been actively contributing to its evolution.

In this project, we will involve members of the vantage6 community to determine what we could create to have the biggest impact on the research community. We will engage this consortium in various brainstorming and feedback sessions.

Participating organisations

Netherlands eScience Center


Related projects


Coronary artery disease: risk estimations and interventions for prevention and Early detection –...

Updated 15 months ago
In progress

Perfect Fit

Targeting key risk factors for cardiovascular disease in at-risk individuals using a personalized and adaptive approach

Updated 9 months ago
In progress


Self tracking for prevention and diagnosis of heart disease

Updated 15 months ago
In progress


Using big data to put a cardiovascular digital twin into the hands of people

Updated 2 months ago
In progress

Related software



An open-source platform for supporting the development of federated learning projects

Updated 3 months ago
2 8