Health monitoring devices are gaining popularity both as wellness tools and as a source of information for healthcare decisions. In this work, we use Spiking Neural Networks (SNNs) for time-series forecasting due to their proven energy-saving capabilities. Thanks to their design that closely mimics the natural nervous system, SNNs are energy-efficient in contrast to classic Artificial Neural Networks (ANNs). We design and implement an energy-efficient privacy-preserving forecasting system on real-world health data streams using SNNs and compare it to a state-of-the-art system with Long short-term memory (LSTM) based prediction model. Our evaluation shows that SNNs tradeoff accuracy (2.2x greater error), to grant a smaller model (19% fewer parameters and 77% less memory consumption) and a 43% less training time. Our model is estimated to consume 3.36 mu J energy, which is significantly less than the traditional ANNs. Finally, we apply epsilon-differential privacy for enhanced privacy guarantees on our federated learning-based models. With differential privacy of epsilon = 0.1, our experiments report an increase in the measured average error (RMSE) of only 25%.

Ensuring user privacy while learning from the acquired Internet of Things sensor data, using limited available compute resources on edge devices, is a challenging task. Ideally, it is desirable to make all the features of the collected data private but due to resource limitations, it is not always possible as it may cause overutilization of resources, which in turn affects the performance of the whole system. In this work, we use the generalization techniques for data anonymization and provide customized injective privacy encoder functions to make data features private. Regardless of the resource availability, some data features must be essentially private. All other data features that may pose low privacy threat are termed as nonessential features. We propose Dynamic Iterative Greedy Search (DIGS), a novel approach with corresponding algorithms to select the set of optimal data features to be private for machine learning applications provided device resource constraints. DIGS selects the necessary and the most private version of data for the application, where all essential and a subset of nonessential features are made private on the edge device without resource overutilization. We have implemented DIGS in Python and evaluated it on Raspberry Pi model A (an edge device with limited resources) for an SVM-based classification on real-life health care data. Our evaluation results show that, while providing the required level of privacy, DIGS allows to achieve up to 26.21% memory, 16.67% CPU instructions, and 30.5% of network bandwidth savings as compared to making all the data private. Moreover, our chosen privacy encoding method has a positive impact on the accuracy of the classification model for our chosen application.

IoT has enabled the creation of a multitude of personal applications and services for a better understanding and improvement of urban environments and our personal lives. These services are driven by the continuous collection and analysis of sensitive and private user data to provide personalised experiences. Among the different application areas of IoT, smart health care, in particular, necessitates the usage of privacy preservation techniques in order to guarantee protection from user privacy-breaching threats such as identification, profiling, localization and tracking, and information linkage. Traditional privacy preservation techniques such as pseudonymization are no longer sufficient to cater to the requirements of privacy preservation in the fast-growing smart health care domain due to the challenges offered by big data volume, velocity, and variety. On the other hand, there is a number of modern privacy preservation techniques with respective overheads that may have a negative impact on application performance such as reduced accuracy, reduced data utility, and increased device resource usage. There is a need to select appropriate privacy preservation techniques (and solutions) according to the nature of data, system performance requirements, and resource constraints, in order to find proper trade-offs between providing privacy preservation, data utility, and acceptable system performance in terms of accuracy, runtime, and resource consumption.

In this work, we investigate different privacy preservation solutions and measure the impact of introducing our selected privacy preservation solutions on the performance of different components of the IoT ecosystem in terms of data utility and system performance. We implement, illustrate, and evaluate the results of our proposed approaches using real-world and synthetic privacy-preserving smart health care datasets. First, we provide a detailed taxonomy and analysis of the privacy preservation techniques and solutions which may serve as a guideline for selecting appropriate techniques according to the nature of data and system requirements. Next, in order to facilitate privacy preserving data sharing, we present and implement a method for creating realistic synthetic and privacy-preserving smart health care datasets using Generative Adversarial Networks and Differential Privacy. Later, we also present and develop a solution for privacy preserving data analytics, a differential privacy library PyDPLib, with health care data as a use case.

In order to find proper trade-offs between providing necessary privacy preservation, device resource consumption, and application accuracy, we present and implement a novel approach with corresponding algorithms and an end-to-end system pipeline for reconfigurable data privacy in machine learning on resource-limited computing devices. Our evaluation results show that, while providing the required level of privacy, our proposed approach allows us to achieve up to 26.21% memory, 16.67% CPU instructions, and 30.5% of network bandwidth savings as compared to making all the data private. Moreover, we also present and implement an end-to-end solution for privacy-preserving time-series forecasting of user health data streams using Federated Learning and Differential Privacy. Our proposed solution finds a proper trade-off between providing necessary privacy preservation, application accuracy, and runtime, and at best introduces a decrease of ~2% in the prediction accuracy of the trained models.

IoT har möjliggjort skapandet av en mängd personliga applikationer och tjänster för en bättre förståelse och förbättring av stadsmiljöer och våra personliga liv. Dessa tjänster drivs av kontinuerlig insamling och analys av känslig och privat användardata för att ge personliga upplevelser. Bland de olika applikationsom- rådena för IoT, kräver i synnerhet smart hälsovård användningen av tekniker för bevarande av integritet för att garantera skydd mot användarnas integritetsintrång, såsom identifiering, profilering, lokalisering och spårning och informationskopp- ling. Traditionella tekniker för bevarande av integritet som pseudonymisering är inte längre tillräckliga för att tillgodose kraven på bevarande av integritet i den snabbväxande smarta hälsovårdsdomänen på grund av de utmaningar som stora datamängder, hastighet och variation forcerar. Å andra sidan finns det ett antal moderna tekniker för bevarande av integritet med respektive omkostnader som kan ha en negativ inverkan på applikationsprestanda såsom minskad noggrannhet, minskad datanytta och ökad resursanvändning på enheten. Det finns ett behov av att välja lämpliga sekretessskyddstekniker (och lösningar) i enlighet med datas natur, systemprestandakrav och resursbegränsningar, för att hitta korrekta avvägning- ar mellan tillhandahållande av integritetsbevarande, dataverktyg och acceptabel systemprestanda i form av av noggrannhet, körtid och resursförbrukning.

I detta arbete undersöker vi olika lösningar för bevarande av integritet och mäter effekten av att introducera våra utvalda lösningar för bevarande av integritet på prestandan hos olika komponenter i IoT-ekosystemet när det gäller datanytta och systemprestanda. Vi implementerar, illustrerar och utvärderar resultaten av våra föreslagna tillvägagångssätt med hjälp av verkliga och syntetiska integritets- bevarande smarta hälsodatauppsättningar. Först tillhandahåller vi en detaljerad taxonomi och analys av tekniker och lösningar för bevarande av integritet som kan fungera som en riktlinje för att välja lämpliga tekniker i enlighet med typen av data och systemkrav. Därefter, för att underlätta integritetsbevarande datadelning, presenterar och implementerar vi en metod för att skapa realistiska syntetiska och integritetsbevarande smarta hälsovårdsdatauppsättningar med hjälp av Ge- nerative Adversarial Networks och Differential Privacy. Senare presenterar och utvecklar vi också en lösning för integritetsbevarande dataanalys, ett differentiellt integritetsbibliotek PyDPLib, med sjukvårdsdata som ett användningsfall.

För att hitta korrekta avvägningar mellan tillhandahållande av nödvändig integri- tetsbevarande, enhetsresursförbrukning och applikationsnoggrannhet presenterar och implementerar vi ett nytt tillvägagångssätt med motsvarande algoritmer och en end-to-end systempipeline för omkonfigurerbar datasekretess i maskininlärning på resursbegränsade datorenheter. Våra utvärderingsresultat visar att, samtidigt som vi tillhandahåller den nödvändiga integritetsnivån, tillåter vårt föreslagna tillvägagångssätt oss att uppnå upp till 26,21% minne, 16,67% CPU-instruktioner och 30,5% av besparingar på nätverkets bandbredd jämfört med att göra all datasammanfattning viiprivat. Dessutom presenterar och implementerar vi också en helhetslösning för integritetsbevarande tidsserieprognoser för användarhälsodataströmmar med hjälp av Federated Learning och Differential Privacy. Vår föreslagna lösning finner en lämplig avvägning mellan att tillhandahålla nödvändig integritetsbevarande, ap- plikationsnoggrannhet och körtid, och introducerar i bästa fall en minskning med ≈ 2% i prediktionsnoggrannheten för de tränade modellerna.

Predictive modeling has the potential to improve risk stratification of cancer patients and thereby contribute to optimized treatment strategies and better outcomes for patients in clinical practice. To develop robust predictive models for decision-making in healthcare, sensitive patient-level data is often required when developing the training models. Consequently, data privacy is an important aspect to consider when building these predictive models and in subsequent communication of the results. In this study we have used Graph Neural Networks for survival prediction, and compared the accuracy to state-of-the-art prediction models after applying Differential Privacy and k-Anonymity, i.e. two privacy-preservation solutions. By using two different data sources we demonstrated that Graph Neural Networks and Survival Forests are the two most well-performing survival prediction methods when used in combination with privacy preservation solutions. Furthermore, when the predictive model was built using clinical expertise in the specific area of interest, the prediction accuracy of the proposed knowledge based graph model drops by at most 10% when used with privacy preservation solutions. Our proposed knowledge based graph is therefore more suitable to be used in combination with privacy preservation solutions as compared to other graph models.

Pharmaceutical and medical technology companies accessing real-world medical data are not interested in personally identifiable data but rather in cohort data such as statistical aggregates, patterns, and trends. These companies cooperate with medical institutions that collect medical data and want to share it but they need to protect the privacy of individuals on the shared data. We present PyDPLib, a Python Differential Privacy library for private medical data analytics. We illustrate an application of differential privacy using PyDPLib in our platform for visualizing private statistics on a database of prostate cancer patients. Our experimental results show that PyDPLib allows creating statistical data plots without compromising patients' privacy while preserving underlying data distributions. Even though PyDPLib has been developed to be used in our platform for reporting the radiological examinations and procedures, it is general enough to be used to provide differential privacy on data in any data analytics and visualization platform, service or application.

With the rapid advancements in machine learning, the health care paradigm is shifting from treatment towards prevention. The smart health care industry relies on the availability of large-scale health datasets in order to benefit from machine learning-based services. As a consequence, preserving the individuals' privacy becomes vital for sharing sensitive personal information. Synthetic datasets with generative models are considered to be one of the most promising solutions for privacy-preserving data sharing. Among the generative models, generative adversarial networks (GANs) have emerged as the most impressive models for synthetic data generation in recent times. However, smart health care data is attributed with unique challenges such as volume, velocity, and various data types and distributions. We propose a GAN coupled with differential privacy mechanisms for generating a realistic and private smart health care dataset. The proposed approach is not only able to generate realistic synthetic data samples but also the differentially private data samples under different settings: learning from a noisy distribution or noising the learned distribution. We tested and evaluated our proposed approach using a real-world Fitbit dataset. Our results indicate that our proposed approach is able to generate quality synthetic and differentially private dataset that preserves the statistical properties of the original dataset.

Privacy preservation plays a vital role in health care applications as the requirements for privacy preservation are very strict in this domain. With the rapid increase in the amount, quality and detail of health data being gathered with smart devices, new mechanisms are required that can cope with the challenges of large scale and real-time processing requirements. Federated learning (FL) is one of the conventional approaches that facilitate the training of AI models without access to the raw data. However, recent studies have shown that FL alone does not guarantee sufficient privacy. Differential privacy (DP) is a well-known approach for privacy guarantees, however, because of the noise addition, DP needs to make a trade-off between privacy and accuracy. In this work, we design and implement an end-to-end pipeline using DP and FL for the first time in the context of health data streams. We propose a clustering mechanism to leverage the similarities between users to improve the prediction accuracy as well as significantly reduce the model training time. Depending on the dataset and features, our predictions are no more than 0.025% far off the ground-truth value with respect to the range of value. Moreover, our clustering mechanism brings a significant reduction in the training time, with up to 49% reduction in prediction accuracy error in the best case, as compared to training a single model on the entire dataset. Our proposed privacy preserving mechanism at best introduces a decrease of ≈ 2% in the prediction accuracy of the trained models. Furthermore, our proposed clustering mechanism reduces the prediction error even in highly noisy settings by as much as 38% as compared to using a single federated private model.

IoT has enabled the creation of a multitude of personal applications and services for a better understanding of urban environments and our personal lives. These services are driven by the continuous collection and analysis of user data in order to provide personalized experiences. However, there is a strong need to address user privacy concerns as most of the collected data is of sensitive nature. This paper provides an overview of privacy preservation techniques and solutions proposed so far in literature along with the IoT levels at which privacy is addressed by each solution as well as their robustness to privacy breaching attacks. An analysis of functional and non-functional limitations of each solution is done, followed by a short survey of machine learning applications designed with these solutions. We identify open issues in the privacy preserving solutions when used in IoT environments. Moreover, we note that most of the privacy preservation solutions need to be adapted in the light of GDPR to accommodate the right to privacy of the users.