Towards a Privacy Compliant Research Interface for Multicenter Medical Data

Big Data analysis gains more and more interest in the processing of e-Health data. The potentially big benefit of those analyses comes with a set of new unknown impacts to an individual's privacy. Still it is important to find a balance between privacy impact and utility of the medical data analysis. To achieve this, this technical report takes a look on different privacy preserving techniques, that could be used for a privacy preserving research interface for medical data. The three techniques Differential privacy, k-Anonymity and Secure multi-party Computation are evaluated on their feasibility for a medical use-case. With those preliminaries some formal definitions are made for a privacy preserving research interface which implements an hybrid approach of the three techniques and a consent based interface.