This demonstration highlights the operational STRATA-FIT vantage6 infrastructure, showcasing real-time data analysis while ensuring stringent data privacy and security standards through federated learning.

Integration of Personal Health Train (PHT) Concepts

In our federated setup:

• Stations, represented by nodes at each participating institution, securely hold sensitive patient data.
• Trains, operationalized as Docker containers, carry specific algorithms to these stations to execute analyses directly on the data.
• Tracks are facilitated by the vantage6 infrastructure, enabling secure algorithmic excursions without relocating the underlying data.

Demonstration Highlights

• Node Control: We illustrate local data sovereignty by starting and stopping a node, demonstrating that when a node is down, it does not establish connections with the server to run tasks, thus emphasizing local control over data accessibility.
• Vantage6 UI Functionality: The demo navigates through the vantage6 UI, showing how to monitor which nodes are active, manage task visualizations, and ensure system integrity.
• Scripting with Vantage6 Client: We demonstrate how to use the vantage6 client within a Jupyter notebook to script and dispatch federated learning tasks, providing a flexible and powerful tool for researchers.
• Results Visualization: The demonstration includes viewing partial and aggregated averages calculated by the algorithm, showcasing the capability of the system to deliver meaningful analytical insights across distributed data sources.
• Data Verification: By executing the demo algorithm, we verify the presence of cohort data at the KI, MUW, and UMCU nodes, confirming that the infrastructure is fully operational and data is accessible for analysis.

Conclusion and Impact

This demonstration confirms the STRATA-FIT project’s capability to conduct sophisticated federated analyses, highlighting the practical application and robustness of the vantage infrastructure in a live environment. By adhering to the Personal Health Train principles, STRATA-FIT enhances collaborative medical research without compromising data privacy.

.

.

Funded by the European Union (grant agreement no. 101080243). Views and opinions expressed are, however, those of the author(s) only and do not necessarily reflect those of the European Union or the European Health and Digital Executive Agency. Neither the European Union nor the granting authority can be held responsible for them.
The project has also received funding from the Swiss State Secretariat for Education, Research and Innovation (SERI) and from Hungary’s National Research, Development and Innovation (NRDI) Fund.