Privacy-Preserving ECG Data Analysis with Differential Privacy: A Literature Review and A Case Study

TL;DR

An overview of key concepts in differential privacy is provided, followed by a literature review and discussion of its application to ECG analysis, and the shortcomings and challenges of applying differential privacy to ECG datasets are discussed.

Abstract

Differential privacy has become the preeminent technique to protect the privacy of individuals in a database while allowing useful results from data analysis to be shared. Notably, it guarantees the amount of privacy loss in the worst-case scenario. Although many theoretical research papers have been published, practical real-life application of differential privacy demands estimating several important parameters without any clear solutions or guidelines. In the first part of the paper, we provide an overview of key concepts in differential privacy, followed by a literature review and discussion of its application to ECG analysis. In the second part of the paper, we explore how to implement differentially private query release on an arrhythmia database using a six-step process. We provide guidelines and discuss the related literature for all the steps involved, such as selection of the $ε$ value, distribution of the total $ε$ budget across the queries, and estimation of the sensitivity for the query functions. At the end, we discuss the shortcomings and challenges of applying differential privacy to ECG datasets.

Figures (10)

• Figure 1: ECG waveform and segments in lead II for normal cardiac cycle
• Figure 2: A twelve-lead ECG showing atrial fibrillation rhythm with no visible P waves that are replaced by coarse fibrillation waves and irregularly irregular QRS complex.
• Figure 3: Steps in a noninteractive differential privacy query release
• Figure 4: Sex distribution in dataset ($\epsilon=0.2$ for all histograms)
• Figure 5: Age distribution in dataset ($\epsilon=0.2$ for all histograms)