Framework for Privacy-Preserving Wearable Health Data Analysis: Proof-of-Concept Study for Atrial Fibrillation Detection

Abstract

Medical wearable devices monitor health data and, coupled with data analytics, cloud computing, and artificial intelligence (AI), enable early detection of disease. Privacy issues arise when personal health information is sent or processed outside the device. We propose a framework that ensures the privacy and integrity of personal medical data while performing AI-based homomorphically encrypted data analytics in the cloud. The main contributions are: (i) a privacy-preserving cloud-based machine learning framework for wearable devices, (ii) CipherML—a library for fast implementation and deployment of deep learning-based solutions on homomorphically encrypted data, and (iii) a proof-of-concept study for atrial fibrillation (AF) detection from electrocardiograms recorded on a wearable device. In the context of AF detection, two approaches are considered: a multi-layer perceptron (MLP) which receives as input the ECG features computed and encrypted on the wearable device, and an end-to-end deep convolutional neural network (1D-CNN), which receives as input the encrypted raw ECG data. The CNN model achieves a lower mean F1-score than the hand-crafted feature-based model. This illustrates the benefit of hand-crafted features over deep convolutional neural networks, especially in a setting with a small training data. Compared to state-of-the-art results, the two privacy-preserving approaches lead, with reasonable computational overhead, to slightly lower, but still similar results: the small performance drop is caused by limitations related to the use of homomorphically encrypted data instead of plaintext data. The findings highlight the potential of the proposed framework to enhance the functionality of wearables through privacy-preserving AI, by providing, within a reasonable amount of time, results equivalent to those achieved without privacy enhancing mechanisms. While the chosen homomorphic encryption scheme prioritizes performance and utility, certain security shortcomings remain open for future development.

Obfuscation Algorithm for Privacy-Preserving Deep Learning-Based Medical Image Analysis

“…For our experiments, the VAE is trained on the Medical MNIST dataset [29]. The dataset contains 6 classes of X-ray images, that are randomly distributed for training (30,000 images) and validation (12,000). More details about the Medical MNIST dataset are presented in Section 2.5.…”

“…However, the changes made to the original HE scheme to allow computations on rational numbers come at a cost in terms of privacy, as it provides lower security than standard schemes. This method was further used in [12] to design a cloud-based platform for deploying ML algorithms for wearable sensor data, focused on data privacy. We have further addressed the security compromise in [13], where we combined a HE scheme based on modulo operations over integers [14], an encoding scheme that enables computations on rational numbers, and a numerical optimization strategy that facilitates training with a fixed number of operations.…”

High Performance Privacy Preserving AI

“…A proof-of-concept study for privacy-preserving atrial fibrillation detection [126] was executed involving the building of a deep learning library for homomorphically encrypted data and then testing it for predicting atrial fibrillation [127]. Atrial fibrillation can be detected by analyzing ECG (electrocardiogram) data that is recorded by wearables and in this work they used the recording from Physionet 2017 challenge [128] as their dataset.…”

Energy Efficient Framework for a AIoT Cardiac Arrhythmia Detection System Wearable during Sport

“…Among many applications, there is a special interest in developing wearable health devices (WHDs) tailored for sports activities [1,2]. These WHDs can help in observing, classifying, and improving an athlete's performance [3], or they can be used to monitor their body response in real time during intense training [4][5][6][7][8][9]. The latter is particularly important to avoid possible injuries or sudden death due to abnormal cardiac activity.…”