Combining data from large groups of patients can lead to new insights and better treatment methods for patients. As a result, TNO and the Netherlands Comprehensive Cancer Organisation (IKNL) are working together to gain insights from data with the aim of reducing the impact of cancer. Likewise, the individual treatment of HIV patients can be improved by learning from the successes and failures of doctors and pharmaceutical companies. But how do we secure sensitive patient data in such a way that patients remain anonymous, privacy is guaranteed and the information avoids falling into the wrong hands? This is possible thanks to Multi-Party Computation (MPC).
Combining different datasets leads to new insights, but also raises privacy issues. Multi-Party Computation, or MPC for short, is a ‘toolbox’ of cryptographic techniques that allow several parties to jointly analyse their data, just as if they have a shared database. Data are protected by cryptographic techniques, allowing them to be shared and analysed without parties ever actually being able to access other parties’ data and without having to make use of a third party to carry out the analyses. With MPC, absolutely no data is disclosed, only combined conclusions and insights are revealed. This can be done in such a way that everyone can verify that the results are correct.
In July 2020, in collaboration with Erasmus MC, Zilveren Kruis and ZorgTTP, TNO demonstrated that it is possible to securely link and analyze personal data from multiple organizations. By using state-of-the-art cryptography, a pilot with synthetic data demonstrated that these types of solutions work in practice. Insurance data can be combined with hospital data in the future to make better predictions about diseases and conditions, such as for heart failure. And those predictions can be used by doctors for personalized health care interventions.
TNO and the Netherlands Comprehensive Cancer Organisation (IKNL) are working together to generate insights from the ever-increasing amounts of data available from cancer patients. This can ultimately help to reduce the impact of cancer by increasing the chances of recovery and preventing cancer in the first place. Within this multidisciplinary collaboration – between doctors, epidemiologists, data scientists and clinical computer scientists, among others – MPC is used to extract more value from the available data of the Netherlands Cancer Registry (NCR) and other sources. The NCR contains data on diseases, care and outcomes.
The ambition of TNO and IKNL for 2020 is to develop privacy-secure, open-source software. This software will be more widely applicable in many more healthcare applications. In this way, these new innovations can be of maximum benefit to the patient.
HIV research also runs into problems regarding the use of privacy-sensitive patient data. MPC technology can offer a solution in this area too. TNO, together with research groups at the University of Amsterdam (UvA) and Centrum Wiskunde & Informatica (CWI), has conducted research into the suitability of MPC when it comes to determining the optimal treatment for an individual HIV patient without compromising privacy.All those involved are enthusiastic and think that MPC can also be used to determine the best treatment for other diseases and within other sectors.
The Care for Data platform was developed in collaboration with CZ, Zuyderland Hospital and CBS. This decentralised platform makes it possible to perform statistical analyses of care data that remain with different parties in a secure manner and of which only the results for the total population are visible (data on a personal level is therefore never shared). Additionally, no Trusted Third Party is needed to anonymise the data.
Care for Data is a scalable platform which makes it possible to continuously assess the effectiveness of care innovations at a low cost or to monitor the quality of care (value-based healthcare) without sensitive data ending up with other parties. TNO is looking for parties that want to evaluate the cost-effectiveness of an existing healthcare innovation or that are interested in monitoring value-based healthcare contracts.
It isn’t just healthcare that benefits from MPC’s unique ability to combine and share data without revealing the content. Through collaboration – read, combining their datasets – financial institutions will also be better able to detect fraud and money laundering activities, for example. Here too, the data that make this possible are sensitive and should not be shared unnecessarily. MPC can improve fraud detection by securely connecting the data.
MPC can also benefit the various governmental parties that work together to trace fugitives who have yet to serve their custodial sentences. Thanks to MPC, parties within the chain of investigation can use one another’s data without sharing sensitive personal data unnecessarily. Want to know how? Download the PDF.
TNO has the unique expertise needed to help your organisation with these techniques and to advise you on customised solutions.
Please contact Daniël Worm
Your question has been sent.
Sorry! Something unexpected happened. Please try again later.