-
Studies in Health Technology and... Jun 2022Substantial advances in methods of collecting and aggregating large amounts of biomedical data have been met with insufficient measures of protecting it from unwarranted... (Review)
Review
Substantial advances in methods of collecting and aggregating large amounts of biomedical data have been met with insufficient measures of protecting it from unwarranted access and use. Most of the current layers of protection are merely aimed at ensuring compliance with regulations (e.g., the EU's General Data Protection Regulation) but do not represent a vision of privacy-by-design as an efficient and ethical advantage in biomedical research and clinical applications. This not only slows down the pace of such efforts but also leaves the data exposed to a wide spectrum of cyberattacks. This work presents an overview of recent advancements in data and compuation security, along with a discussion of their limitations and potential for deployement in both health care and research settings.
Topics: Biomedical Research; Computer Security; Confidentiality; Privacy
PubMed: 35673008
DOI: 10.3233/SHTI220069 -
The Veterinary Record Sep 2018
Topics: Computer Security; Humans; Legislation, Veterinary; Privacy; Publishing; Societies; United Kingdom; Veterinary Medicine
PubMed: 30266857
DOI: 10.1136/vr.k3956 -
Neural Networks : the Official Journal... Nov 2022A feedforward-designed convolutional neural network (FF-CNN) is an interpretable neural network with low training complexity. Unlike a neural network trained using...
A feedforward-designed convolutional neural network (FF-CNN) is an interpretable neural network with low training complexity. Unlike a neural network trained using backpropagation (BP) algorithms and optimizers (e.g., stochastic gradient descent (SGD) and Adam), a FF-CNN obtains the model parameters in one feed-forward calculation based on two methods of data statistics: subspace approximation with adjusted bias and least squares regression. Currently, models based on FF-CNN training methods have achieved outstanding performance in the fields of image classification and point cloud data processing. In this study, we analyze and verify that there is a risk of user privacy leakage during the training process of FF-CNN and existing privacy-preserving methods for model gradients or loss functions do not apply to FF-CNN models. Therefore, we propose a securely forward-designed convolutional neural network algorithm (SFF-CNN) to protect the privacy and security of data providers for the FF-CNN model. Firstly, we propose the DPSaab algorithm to add the corresponding noise to the one-stage Saab transform in the FF-CNN design for improved protection performance. Secondly, because noise addition brings the risk of model over-fitting and further increases the possibility of privacy leakage, we propose the SJS algorithm to filter the input features of the fully connected model layer. Finally, we theoretically prove that the proposed algorithm satisfies differential privacy and experimentally demonstrate that the proposed algorithm has strong privacy protection. The proposed algorithm outperforms the compared deep learning privacy-preserving algorithms in terms of utility and robustness.
Topics: Privacy; Neural Networks, Computer; Algorithms; Cloud Computing
PubMed: 36027662
DOI: 10.1016/j.neunet.2022.08.005 -
Travel Medicine and Infectious Disease 2021The advent of mobile applications for health and medicine will revolutionize travel medicine. Despite their many benefits, such as access to real-time data, mobile apps... (Review)
Review
BACKGROUND
The advent of mobile applications for health and medicine will revolutionize travel medicine. Despite their many benefits, such as access to real-time data, mobile apps for travel medicine are accompanied by many ethical issues, including questions about security and privacy.
METHODS
A systematic literature review as conducted following PRISMA guidelines. Database screening yielded 1795 results and seven papers satisfied the criteria for inclusion. Through a mix of inductive and deductive data extraction, this systematic review examined both the benefits and challenges, as well as ethical considerations, of mobile apps for travel medicine.
RESULTS
Ethical considerations were discussed with varying depth across the included articles, with privacy and data protection mentioned most frequently, highlighting concerns over sensitive information and a lack of guidelines in the digital sphere. Additionally, technical concerns about data quality and bias were predominant issues for researchers and developers alike. Some ethical issues were not discussed at all, including equity, and user involvement.
CONCLUSION
This paper highlights the scarcity of discussion around ethical issues. Both researchers and developers need to better integrate ethical reflection at each step of the development and use of health apps. More effective oversight mechanisms and clearer ethical guidance are needed to guide the stakeholders in this endeavour.
Topics: Humans; Mobile Applications; Privacy; Travel Medicine
PubMed: 34256131
DOI: 10.1016/j.tmaid.2021.102143 -
Journal of Nursing Management Nov 2022To explore current research on the ethics of smart home technologies including artificial intelligence and information technologies for elderly care by conducting a... (Review)
Review
AIM
To explore current research on the ethics of smart home technologies including artificial intelligence and information technologies for elderly care by conducting a scoping review.
BACKGROUND
The development of smart home technologies for care of the older adults provides potential solutions to reduce the caregiver burden within families where they are urgently needed. Building an ethical system to support the application of these technical products should be explored.
METHODS
The literature search was performed in seven electronic databases. Relevant studies from January 2015 to February 2021 were selected; screening and analysis were completed independently by two researchers.
RESULTS
There were a total of 15 included studies on the ethics of smart home technologies for elderly care, which focused on the following issues: privacy (information privacy and physical privacy), autonomy (independence, informed consent and user-centred control), safety guarantee, fairness and concerns about reduced human contact.
CONCLUSIONS
There exist a number of ethical conflicts in the application of smart home technologies for elderly care. Therefore, it is necessary to further investigate the ethical issues with regards to the decision-making process of weighing the advantages and disadvantages of these technologies.
IMPLICATIONS FOR NURSING MANAGEMENT
Efforts should be made to establish a corresponding ethical framework to ensure the sustainable development of smart, home-based elderly care. Nurses may play an important role in the design and implementation of these technologies to promote ethical awareness and practice.
Topics: Humans; Aged; Artificial Intelligence; Privacy
PubMed: 34806243
DOI: 10.1111/jonm.13521 -
Drug Discovery Today Dec 2023Data availability, data security, and privacy concerns often hamper optimal performance efficiency of machine learning (ML) techniques. Therefore, novel techniques for... (Review)
Review
Data availability, data security, and privacy concerns often hamper optimal performance efficiency of machine learning (ML) techniques. Therefore, novel techniques for the utilization of private/sensitive data in the field of drug discovery have been proposed for ML model-building tasks. Some examples of the different techniques are secure multiparty computation, distributed deep learning, homomorphic encryption, blockchain-based peer-to-peer networking, differential privacy, and federated learning, as well as combinations of such techniques. In this paper, we present an overview of these techniques for decentralized ML to illustrate its benefits and drawbacks in the field of drug discovery.
Topics: Privacy; Drug Discovery; Machine Learning
PubMed: 37935330
DOI: 10.1016/j.drudis.2023.103820 -
Health Care Analysis : HCA : Journal of... Mar 2022Information is clearly vital to public health, but the acquisition and use of public health data elicit serious privacy concerns. One strategy for navigating this...
Information is clearly vital to public health, but the acquisition and use of public health data elicit serious privacy concerns. One strategy for navigating this dilemma is to build 'trust' in institutions responsible for health information, thereby reducing privacy concerns and increasing willingness to contribute personal data. This strategy, as currently presented in public health literature, has serious shortcomings. But it can be augmented by appealing to the philosophical analysis of the concept of trust. Philosophers distinguish trust and trustworthiness from cognate attitudes, such as confident reliance. Central to this is value congruence: trust is grounded in the perception of shared values. So, the way to build trust in institutions responsible for health data is for those institutions to develop and display values shared by the public. We defend this approach from objections, such as that trust is an interpersonal attitude inappropriate to the way people relate to organisations. The paper then moves on to the practical application of our strategy. Trust and trustworthiness can reduce privacy concerns and increase willingness to share health data, notably, in the context of internal and external threats to data privacy. We end by appealing for the sort of empirical work our proposal requires.
Topics: Attitude; Humans; Privacy; Public Health; Trust
PubMed: 34751865
DOI: 10.1007/s10728-021-00436-y -
Sensors (Basel, Switzerland) Apr 2022Most traditional agricultural traceability systems are centralized, which could result in the low reliability of traceability results, enterprise privacy data leakage...
Most traditional agricultural traceability systems are centralized, which could result in the low reliability of traceability results, enterprise privacy data leakage vulnerabilities, and the generation of information islands. To solve the above problems, we propose a trusted agricultural product traceability system based on the Ethereum blockchain in this paper. We designed a dual storage model of "Blockchain+IPFS (InterPlanetary File System)" to reduce the storage pressure of the blockchain and realize efficient information queries. Additionally, we propose a data privacy protection solution based on some cryptographic primitives and the Merkle Tree that can avoid enterprise privacy and sensitive data leakage. Furthermore, we implemented the proposed system using the Ethereum blockchain platform and provided the cost, performance, and security analysis, as well as compared it with the existing solutions. The results showed that the proposed system is both efficient and feasible and can meet the practical application requirements.
Topics: Blockchain; Computer Security; Privacy; Reproducibility of Results
PubMed: 35591077
DOI: 10.3390/s22093388 -
Science (New York, N.Y.) Aug 2018
Topics: Genealogy and Heraldry; Genetic Privacy; Genetic Testing; Pedigree; Privacy
PubMed: 30166479
DOI: 10.1126/science.aav0330 -
Cyberpsychology, Behavior and Social... Jul 2018The study contributes to the ongoing debate about the "privacy paradox" in the context of using social media. The presence of a privacy paradox is often declared if...
The study contributes to the ongoing debate about the "privacy paradox" in the context of using social media. The presence of a privacy paradox is often declared if there is no relationship between users' information privacy concerns and their online self-disclosure. However, prior research has produced conflicting results. The novel contribution of this study is that we consider public and private self-disclosure separately. The data came from a cross-national survey of 1,500 Canadians. For the purposes of the study, we only examined the subset of 545 people who had at least one public account and one private account. Going beyond a single view of self-disclosure, we captured five dimensions of self-disclosure: Amount, Depth, Polarity, Accuracy, and Intent; and two aspects of privacy concerns: concerns about organizational and social threats. To examine the collected data, we used Partial Least Squares Structural Equation Modeling. Our research does not support the presence of a privacy paradox as we found a relationship between privacy concerns from organizational and social threats and most of the dimensions of self-disclosure (even if the relationship was weak). There was no difference between patterns of self-disclosure on private versus public accounts. Different privacy concerns may trigger different privacy protection responses and, thus, may interact with self-disclosure differently. Concerns about organizational threats increase awareness and accuracy while reducing amount and depth, while concerns about social threats reduce accuracy and awareness while increasing amount and depth.
Topics: Adult; Canada; Confidentiality; Female; Humans; Intention; Male; Middle Aged; Privacy; Self Disclosure; Social Media; Young Adult
PubMed: 29995525
DOI: 10.1089/cyber.2017.0709