gms | German Medical Science

Introduction: Data-driven medical research has the potential to generate insights that can significantly improve prevention, diagnostics, and therapy [1]. One way to obtain the necessary data is through data sharing - a practice in which various medical institutions share data with researchers. However, given the sensitivity of biomedical data, data sharing is constrained by laws such as the European Union (EU) General Data Protection Regulation (GDPR). To address the challenge of data sharing while maintaining privacy, several approaches have been proposed. One of them is Secure Multi-Party Computation (SMPC), a cryptography-based set of methods that allow mutual computation of a common result between parties without sharing their input data. Although SMPC is very promising, it is rarely used in practice today. Two major problems associated with the practical application of SMPC are its technical complexity and the lack of usability of many SMPC-based tools.

Methods: To address these challenges, we have developed the user-friendly tool EasySMPC [2]. In this poster, we briefly introduce EasySMPC and address recent developments and future challenges. EasySMPC uses the GMW protocol [3], allowing the computation of common sums of variables from different parties without revealing the summands. The communication layer is kept flexible to allow deployment without additional server components. In its original form, the software supports automated data exchange via emails. In order to enable larger bandwidths and more stable connections, further forms of communication have been implemented recently and a command line interface (CLI) has been developed for automated use in processing pipelines.

Results: Important recent developments and extensions of EasySMPC include two different communication modes. First, a dedicated backend, called EasyBackend, has been developed as a RESTful service and a central instance has been deployed.. Second, an interface is being developed against Samply.Beam, a software component to enable fast and secure communication across sites. The developed CLI supports all features that are also accessible through the graphical application. In future work, we aim to extend EasySMPC in the direction of a spreadsheet application allowing simple use and combination of different SMPC-methods. Supporting more such functions will also require extending EasySMPC with additional protocols that support further arithmetic operations.

Discussion: EasySMPC is an easy-to-use tool aiming for non-technical experts as users and it has been designed to popularize the use of SMPC protocols in practice. Due to its simplicity, the tool is also suitable for answering open questions about SMPC technologies at the intersection of law and technology on the basis of an exemplary application. For example, CORD_MI [4] is currently working on a real-world deployment to calculate statistics about comorbidities of patients with rare diseases across several university hospitals. The corresponding coordination with and evaluation of compliance experts is facilitated by the reduction to the essential in the software.

Conclusion: Development of EasySMPC is progressing with the aim to foster the adoption of SMPC technologies. Future work will focus on extending the statistical capabilities of the software. EasySMPC is licensed under the permissive Apache License 2.0.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.