Graduate School and Research Center in Digital Sciences

PAC: Privacy-preserving arrhythmia classification with neural networks

Mansouri, Mohamad; Bozdemir, Beyza; Önen, Melek; Ermis, Orhan

Technical report RR-19-341, 30 August 2019

Recent advances in information technology such as the Internet of Things enable businesses and organisations to collect large amounts of data and apply advanced machine learning techniques in order to infer valuable insights and improve predictions. Unfortunately, such bene ts come with a high cost in terms of privacy exposures given the high sensitivity of the data that are usually analysed/processed at third party servers. In this study, we aim at protecting health data, more precisely, Electro-Cardiogram (ECG) data while enabling an effi cient prediction of arrhythmia using neural networks. We propose a solution named PAC that combines the use of Neural Networks with secure two-party computation. To achieve a good trade-off between privacy, accuracy, and efficiency, we rst build a dedicated, efficient neural network model for which the underlying operations can be further performed while data is privacy protected. The resulting model consists of two fully connected layers which perform small size matrix multiplication and one activation layer which executes a square function. The solution is implemented using the ABY framework and makes use of Arithmetic and Boolean circuits. Several approximations are performed over the representation of the data, accordingly. PAC also supports classi cations in batches when needed. Experimental results run on the PhysioBank datasets show an accuracy of 96.34% which outperforms existing solutions.

Document Bibtex

Title:PAC: Privacy-preserving arrhythmia classification with neural networks
Keywords:data privacy, privacy-preserving neural networks, secure two-party computation, arrhythmia classi cation
Department:Digital Security
Eurecom ref:5998
Copyright: © EURECOM. Personal use of this material is permitted. The definitive version of this paper was published in Technical report RR-19-341, 30 August 2019 and is available at :
Bibtex: @techreport{EURECOM+5998, year = {2019}, title = {{PAC}: {P}rivacy-preserving arrhythmia classification with neural networks}, author = {{M}ansouri, {M}ohamad and {B}ozdemir, {B}eyza and {\"{O}}nen, {M}elek and {E}rmis, {O}rhan}, number = {EURECOM+5998}, month = {08}, institution = {Eurecom}, url = {},, }
See also: