EPI - Enabling Personalized Interventions
In this project a platform for the safe and reliable sharing of large heterogeneous data sources and data analysis techniques will be developed. The aim is to determine the best medical and lifestyle interventions given the relevant personal health situation of users of this platform.
In order to be able to give people such health advice attributed to their personal situation, the data of these patients must be analysed as well as possible. Different data sources will have to be made available for this purpose. The privacy-sensitivity of medical data and restrictions on access to and processing of this data place high demands on the data infrastructure. The data governance of such a data infrastructure must enable the data subjects (often: patients) to give, withdraw and enforce permission to use, that only permitted processing of this data can take place (embedded compliance) and that the use of this data is administered in such a way that its legitimacy can be guaranteed.
Data protection regulations mean that the data processing algorithms (machine learning, statistics), which sometimes also have (copyright) protection, must be distributed over the various data collections. Such a distributed solution is new and raises interesting questions, including those relating to:
(a) the effects on the learning outcomes (how much data is needed to reach a certain conclusion about whether or not a certain lifestyle advice works?),
(b) the scalability of the solution and,
(c) the integrability of the links learned with the system that informs users of the health advice tailored to their personal situation.
For example, the data was obtained partly from sensors, partly from the general practitioner, partly from the hospital, partly from a questionnaire filled in by means of an app. On the basis of this data, advice is given, which in turn generates new data. This data is also made available for further analysis, provided that permission has been given, which results in a health ecosystem that can be used to provide better and better advice. The required data analysis requires new machine learning and statistical techniques, whereby patients must be dynamically clustered into small groups, and the cluster composition may change over time.
This project is carried out by an interdisciplinary team of medical specialists, data scientists, ICT infrastructure experts and experts in the field of artificial intelligence and law, which includes all the relevant disciplines required to arrive at an adequate solution. The project will be carried out on the basis of a number of representative health-related cases, involving not only the healthcare institutions involved but also a leading supplier of diagnostic equipment. The results will therefore already have an impact during the implementation of the project and thus help people to improve their health situation.
Sign up for our newsletter and stay informed of the latest news Commit2data