MAster assignment
Semantic support for the personal health train
Period: (TBD)
Student: Unassigned
If you are interested please contact:
Traditional approaches to data analysis require the data to be available at the location where the analysis processing takes place. These approaches have worked well with limited and centralised data but are problematic when the data to be processed are scattered and/or are subject to privacy requirements concerning their accessibility. In order to address this problem, federated (distributed) schemes for executing processes on data have been developed, including the Personal Health Train (PHT), which has been developed to regulate the access to personal health data. There are already some PHT prototypes that handle incoming trains (processing units) at a location where data are available (a data station), and that are capable of staging the data station in the cloud, if not enough resources are available at the data station. However, many aspects of the PHT approach need to be further investigated and developed. Some examples of research questions are:
- How to allow trains to combine data available at multiple data stations?
- How to properly describe the data stations and trains (their metadata) so that reasoning about them is possible? How to implement this reasoning?
- What are the proper APIs necessary to make this approach modular, interoperable and reusable?
- Which case studies can be defined to validate and help improve the PHT approach?
- How to define the train dispatching planning based on the required data and the stations that need to be visited?
Each of these questions can be addressed in a different Master project. If possible, we want to have multiple students working at the same time on these different questions, so that they can help each other.
These projects are part of our collaboration with the Leiden University Medical Center (LUMC), so researchers from LUMC may be involved in the supervision of these projects.
Keywords: distributed data analysis, Personal Health Train, FAIR data stations and trains, data privacy preservation