Fall Focused Research Program Workshop: Secure analytics on mobile health data streams

Date: Thursday, November 18th, 2021

Time: 2:00pm – 5:00pm ET

Location (in-person & zoom): Rajen Kilachand Center for Integrated Life Sciences and Engineering (KCILSE), 610 Commonwealth Avenue, Conference Room 106C, Boston, MA 02215. See workshop webpage for zoom meeting details.

Workshop Overview:

The workshop will bring together researchers from BU’s Medical Campus, BU Computer Science and Engineering, and external collaborators and will include talks on streaming data processing and secure data analytics in the Cloud. We aim to discuss the requirements for designing an end-to-end platform that collects patient data in a streaming fashion (e.g. from smartphones, sensors, wearable devices, etc.) and performs cloud-based analytics with zero information leakage to the cloud provider or any other unauthorized entities.

The workshop is open to interested faculty members, staff and students. If you are interested in participating, please join us! If you have questions or would like more information about this FRP, please contact Xiaoyu Li (lxy2020@bu.edu) for more information.

This workshop will be in-person and held via Zoom Meeting upon request.

Agenda

2:00 – 2:10pm Welcome & about the workshop & FRP

2:10 – 2:30pm Introduction to Stream Processing (Vasiliki Kalavri, Ph.D)

Abstract: Stream processing enables the analysis of continuous, possibly unbounded datasets and facilitates real-time applications by shortening the lag between receiving information and acting on it. In this talk, I will cover the fundamentals of streaming computations, discuss emerging applications, and give an overview of the capabilities of modern stream processing systems.

2:30 – 3:00pm Cryptography and the Power of Learning Nothing (Mayank Varia, Ph.D)

Abstract: In a data-driven society, cryptography can lower the barriers to data science by allowing analysts to work with data that they cannot see. In this talk, I will describe our work to design and deploy systems that compute analytics while respecting personal privacy.

3:00 – 3:20pm Secure collaborative analytics with Secrecy (John Liagouris, Ph.D)

Abstract: Secrecy is a new relational framework for secure collaborative analytics that allows composing and executing end-to-end oblivious queries under secure multi-party computation (MPC). Secrecy’s core novelty is a set of logical and physical optimizations that effectively reduce query execution costs while retaining the full security guarantees of MPC. In this talk, I will give an overview of the Secrecy framework and discuss our plans for extending it to support use-cases in continuous digital health analytics.

3:20 – 3:30pm: Coffee Break

3:30 – 4:00pm: Secure multi-party computation to count patients in a federated clinical data network & Demo of the SHRINE project (Griffin Weber, M.D., Ph.D. & William Simons, Harvard Medical School)

Abstract: The Shared Health Research Information Network (SHRINE) is an open source federated network that enables scientists to run aggregate count queries on the clinical data of more than 100 million patients at dozens of hospitals across the United States. It is used for many clinical studies, including large COVID-19 related research projects. To protect privacy when only a small number of patients match a query, we created a multi-party computation (MPC) prototype that shows the total number of patients in the network without revealing the hospitals where they received care.

4:00 – 4:15pm: Open discussion: Query interfaces

4:15 – 4:40pm: RPis & Secrecy Demo (Muhammad Faisal and Eric Chen)

4:40pm – 4:55pm: Moderated discussion

4:55 – 5:00pm: Closing remarks

Registration

We request all participants to register for the FRP Workshop by 11/16. Please register by completing the survey below, and selecting “submit.”