The use of personal and personally identifiable information, which also and especially include biometric data, is strictly regulated in the context of the General Data Protection Regulation (GDPR). Patients and users in Europe enjoy extensive legal protection against unauthorised use and disclosure of their data. In many cases, however, this strong protection also results in a data usability that is restricted to a greater or lesser degree, which in turn conflicts with its social value for the advancement of medical knowledge and the further development of technologies. Addressing and resolving this dilemma calls for technical solutions that enable a provision and use of data that comply with data protection regulations in the sense of open data so that requirements regarding data protection and data use can be addressed simultaneously.
Objectives and approach
In the »NEMO« project, experts are examining the extent to which a person can be clearly identified from recorded biosignals. The question then is how suitable anonymisation techniques can prevent the identification and disclosure of sensitive information without stripping the data of their scientific value.
The key objective of »NEMO« is to explore and validate innovative techniques for the re-identification analysis and adaptive anonymisation of biosignals, using the example of electroencephalograms (EEG) from sleep monitoring systems. An EEG records brain activity via electrodes attached to the head. On the one hand, these data are particularly sensitive due to their information density, and on the other hand they are also particularly challenging in terms of developing effective anonymisation techniques. At the same time, the number of products on the market that record EEG data in the consumer sector is increasing.
Within »NEMO«, the experts are first of all developing analysis methods to quantify the risks of disclosing identities and sensitive information in the recorded raw data. To minimise these risks and at the same time ensure the highest possible utility of the EEG data, they will then explore and test techniques for their adaptive anonymisation. Among other things, the basis for this is know-how in the anonymisation of audio data and other technical data protection methods.
To integrate users in the anonymisation process and explain to them the mode of action and the added value of the anonymisation variants deployed, a demonstrator is additionally being developed that enables an application-related exploration and analysis of EEG data with the help of the anonymisation techniques made available.
Innovations and outlook
The project aims to generate a significant knowledge gain, not only by illustrating specific risk scenarios but also by developing and testing anonymisation techniques in the sensitive area of health data. The intention is to create a sound foundation for data protection concepts and anonymisation techniques. On the basis of a tried and tested technical infrastructure, the goal is the comprehensive use of data in research and development while at the same time safeguarding data protection.