Building a common and privacy-preserving front end for open-source clinical research platforms

Mickaël Misbach
2018
Student project

Abstract

Being able to exploit large and heterogeneous medical data is crucial for realizing the promise of precision medicine to its full potential. Yet, currently, due to the presence of multiple and fragmented systems at different clinical sites, it is often difficult to enable researchers to access the data they need. Privacy and security concerns also represent major obstacles that need to be overcome in order to provide access to sensitive medical data that are usually not exposed by clinical sites for the fear of data leaks. To address these challenges, we propose a system that employs Glowing Bear as a common front end to the widespread clinical research platforms i2b2, tranSMART and MedCo. Our proposed system uses IRCT, the official implementation of the PIC-SURE API, which acts as an interoperability layer that translates clinical research queries into native API languages. With our work, we take a step towards the technical convergence of i2b2 and tranSMART, giving clinical sites the means to share their data more easily. Even more, with the support of MedCo, a cohort explorer with strong privacy and security guarantees based on federated i2b2 instances, we enable clinical sites to share sensitive data that would be difficult to share otherwise.

Official source

https://infoscience.epfl.ch/record/283388?ln=en

About this result

This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.

Related concepts

Related publications

MedCo: Enabling Secure and Privacy-Preserving Exploration of Distributed Clinical and Genomic Data

Bryan Alexander Ford, Joao André Gomes de Sá E Sousa, Jean-Pierre Hubaux, Olivier Michielin, Mickaël Misbach, Jean Louis Raisaro, Juan Ramón Troncoso-Pastoriza

The increasing number of health-data breaches is creating a complicated environment for medical-data sharing and, consequently, for medical progress. Therefore, the development of new solutions that can reassure clinical sites by enabling privacy-preserving sharing of sensitive medical data in compliance with stringent regulations (e.g., HIPAA, GDPR) is now more urgent than ever. In this work, we introduce MedCo, the first operational system that enables a group of clinical sites to federate and collectively protect their data in order to share them with external investigators without worrying about security and privacy concerns. MedCo uses (a) collective homomorphic encryption to provide trust decentralization and end-to-end confidentiality protection, and (b) obfuscation techniques to achieve formal notions of privacy, such as differential privacy. A critical feature of MedCo is that it is fully integrated within the i2b2 (Informatics for Integrating Biology and the Bedside) framework, currently used in more than 300 hospitals worldwide. Therefore, it is easily adoptable by clinical sites. We demonstrate MedCo’s practicality by testing it on data from The Cancer Genome Atlas in a simulated network of three institutions. Its performance is comparable to the ones of SHRINE (networked i2b2), which, in contrast, does not provide any data protection guarantee.

2019

Most communication systems (e.g., e-mails, instant messengers, VPNs) use encryption to prevent third parties from learning sensitive information.However, encrypted communications protect the contents but often leak metadata: the amount of data sent and the time it was sent, the way the data should be decrypted, the identity of the sender and the recipient.These metadata are a pervasive threat to privacy: They enable a variety of attacks that range from recovering plaintext contents from encrypted communications to inferring communicating parties.Our goal in this thesis is two-fold: First, to raise awareness about this problem by demonstrating attacks that threaten user privacy; and second, to propose novel solutions that reduce the metadata leakage and maintain acceptable usability and costs.To achieve the first goal, we present the first work that performs an in-depth analysis of the communications of wearable devices under the lens of traffic analysis.We demonstrate that the metadata of Bluetooth communications of wearable devices (smartwatches, fitness trackers, and blood-pressure monitors) leak sensitive information to a passive observer, despite the use of encryption.By design, these devices handle fine-grained and long-term, personal, medical and lifestyle-related data from their users.Our work highlights the need to rethink how sensitive data is exchanged in this setting.More generally, we confirm that metadata can pose a threat to user privacy, even in settings where traffic-analysis attacks are perhaps not an immediate threat.For the second goal, we begin by presenting a theoretical contribution that concerns ciphertext formats:Padmé, a padding function designed to reduce the metadata leakage of files and messages through their length.Padmé efficiently hides the size of objects even when they have very different sizes.Padmé is a part of PURBs, a ciphertext format that does not leak metadata, except for a small amount about the size.Then, we design systems that enable communicating while reducing metadata leakage: Anonymous Communication Networks (ACNs).ACNs protect against some metadata leakage (e.g., who is communicating and when).Traffic-analysis resistant ACNs have not seen widespread adoption yet, possibly due to their technical shortcomings.We focus on two aspects that remain unsolved: small-scale traffic-agnostic communication that achieves low latency, and large-scale asynchronous messaging that handles millions of users.We present PriFi, an ACN that provides provable traffic-analysis resistance and low-latency in the context of a local-area network.The protocol hides the source and recipient of a message by ensuring that the communication patterns of all participants are equal, even in the presence of active attacks.PriFi ensures low-latency and traffic-agnostic communication for a small set of users.We then present Rubato, an ACN for anonymous messaging that handles millions of users.Other large-scale ACNs have an important limitation: A sender and a recipient need to be online at the same time to resist intersection attacks.As participants are uncoordinated, they need to be online and to send cover traffic at all times, which is infeasible for many users.Rubato is a circuit-based mixnet that enables its users to be asynchronous; they can participate in the network according to their own schedule.Unlike previous ACNs, this enables users to participate using their mobile devices.

EPFL2021