A5046241222- Privacy-Preserving Technologies in Data
- Ethics in Clinical Research
- Data-Driven Disease Surveillance
- COVID-19 Digital Contact Tracing
- Scientific Computing and Data Management
- Research Data Management Practices
- Privacy, Security, and Data Protection
- Mobile Health and mHealth Applications
- Cell Image Analysis Techniques
- Data Quality and Management
- Health, Environment, Cognitive Aging
- Patient Dignity and Privacy
- Census and Population Estimation
- Neurobiology and Insect Physiology Research
- Ethics and Social Impacts of AI
- Insect and Arachnid Ecology and Behavior
- Health Systems, Economic Evaluations, Quality of Life
- Health disparities and outcomes
- Neuroethics, Human Enhancement, Biomedical Innovations
- Digital Mental Health Interventions
- demographic modeling and climate adaptation
- Advanced Biosensing Techniques and Applications
- Mobile Crowdsensing and Crowdsourcing
- Insurance, Mortality, Demography, Risk Management
- Electronic Health Records Systems
Berlin Institute of Health at Charité - Universitätsmedizin Berlin
2020-2025
Freie Universität Berlin
2020-2021
Humboldt-Universität zu Berlin
2020
Charité - Universitätsmedizin Berlin
2020
Abstract Background Data sharing is considered a crucial part of modern medical research. Unfortunately, despite its advantages, it often faces obstacles, especially data privacy challenges. As result, various approaches and infrastructures have been developed that aim to ensure patients research participants remain anonymous when shared. However, protection typically comes at cost, e.g. restrictions regarding the types analyses can be performed on shared data. What lacking systematization...
Abstract The Lean European Open Survey on SARS-CoV-2 Infected Patients (LEOSS) is a registry for studying the epidemiology and clinical course of COVID-19. To support evidence-generation at rapid pace required in pandemic, LEOSS follows an Science approach, making data available to public real-time. protect patient privacy, quantitative anonymization procedures are used continuously published stream consisting 16 variables therapy COVID-19 from singling out, inference linkage attacks. We...
Abstract Background Sharing health data holds great potential for advancing medical research but also poses many challenges, including the need to protect people’s privacy. One approach address this is anonymization, which refers process of altering or transforming a dataset preserve privacy individuals contributing data. To this, models have been designed measure risks and optimization algorithms can be used transform achieve good balance between reduction preservation dataset’s utility....
The novel coronavirus SARS-CoV-2 rapidly spread around the world, causing disease COVID-19. To contain virus, much hope is placed on participatory surveillance using mobile apps, such as automated digital contact tracing, but broad adoption an important prerequisite for associated interventions to be effective. Data protection aspects are a critical factor adoption, and privacy risks of solutions developed often need balanced against their functionalities. This reflected by intensive...
Background Data provenance refers to the origin, processing, and movement of data. Reliable precise knowledge about data has great potential improve reproducibility as well quality in biomedical research and, therefore, foster good scientific practice. However, despite increasing interest on technologies literature their implementation other disciplines, these have not yet been widely adopted research. Objective The aim this scoping review was provide a structured overview body methods by...
Background Sharing data from clinical studies can accelerate scientific progress, improve transparency, and increase the potential for innovation collaboration. However, privacy concerns remain a barrier to sharing. Certain concerns, such as reidentification risk, be addressed through application of anonymization algorithms, whereby are altered so that it is no longer reasonably related person. Yet, alterations have influence set’s statistical properties, privacy-utility trade-off must...
The identification of vulnerable records (targets) is an important step for many privacy attacks on protected health data. We implemented and evaluated three outlier metrics detecting potential targets. Next, we assessed differences similarities between the top-k targets suggested by different methods studied how susceptible those are to membership inference synthetic Our results suggest that there no one-size-fits-all approach target selection should be chosen based type attack performed.
Abstract Background Data anonymization is an important building block for ensuring privacy and fosters the reuse of data. However, transforming data in a way that preserves subjects while maintaining high degree quality challenging particularly difficult when processing complex datasets contain number attributes. In this article we present how extended open source software ARX to improve its support high-dimensional, biomedical datasets. Findings For improving ARX's capability find optimal...
Navigating animals combine multiple perceptual faculties, learn during exploration, retrieve multi-facetted memory contents, and exhibit goal-directedness as an expression of their current needs motivations. Navigation in insects has been linked to a variety underlying strategies such path integration, view familiarity, visual beaconing, goal-directed orientation with respect previously learned ground structures. Most works, however, study navigation either from field perspective, analyzing...
Abstract Anonymization has the potential to foster sharing of medical data. State-of-the-art methods use mathematical models modify data reduce privacy risks. However, degree protection must be balanced against impact on statistical properties. We studied an extreme case this trade-off: validity open dataset based German National Pandemic Cohort Network (NAPKON), which was prepared for publication using a strong anonymization procedure. Descriptive statistics and results regression analyses...
Sharing biomedical data for research can help to improve disease understanding and support the development of preventive, diagnostic, therapeutic methods. However, it is vital balance amount shared sharing mechanism chosen with privacy protection provided. This requires a detailed potential adversaries who might attempt re-identify consequences their actions. The aim this paper present comprehensive list types adversaries, motivations, harms targeted individuals. A group 13 researchers...
The SARS-CoV-2 pandemic highlighted the importance of fast, collaborative research in biomedicine. Within ORCHESTRA consortium, we rapidly deployed a pseudonymization service with minimal training and maintenance efforts under time-critical conditions to support complex, multi-site project. Over two years, was 13 sites across 11 countries register more than 10,000 study participants 15,000 biosamples. In this work, present lessons learned as part process. Most importantly, that common...
Introduction: A modern approach to ensuring privacy when sharing datasets is the use of synthetic data generation methods, which often claim outperform classic anonymization techniques in trade-off between utility and privacy. Recently, it was demonstrated that various deep learning-based approaches are able generate useful synthesized datasets, based on domain-specific analyses. However, evaluating implications releasing remains a challenging problem, especially goal conform with protection...
<sec> <title>BACKGROUND</title> The novel coronavirus SARS-CoV-2 rapidly spread around the world, causing disease COVID-19. To contain virus, much hope is placed on participatory surveillance using mobile apps, such as automated digital contact tracing, but broad adoption an important prerequisite for associated interventions to be effective. Data protection aspects are a critical factor adoption, and privacy risks of solutions developed often need balanced against their functionalities....
<sec> <title>BACKGROUND</title> Sharing data from clinical studies can accelerate scientific progress, improve transparency, and increase the potential for innovation collaboration. However, privacy concerns remain a barrier to sharing. Certain concerns, such as reidentification risk, be addressed through application of anonymization algorithms, whereby are altered so that it is no longer reasonably related person. Yet, alterations have influence set’s statistical properties, privacy-utility...
In this study, we propose a unified evaluation framework for systematically assessing the utility-privacy trade-off of synthetic data generation (SDG) models. These SDG models are adapted to deal with longitudinal or tabular stemming from electronic health records (EHR) containing both discrete and numeric features. Our considers different sharing scenarios attacker
<sec> <title>BACKGROUND</title> Data provenance refers to the origin, processing, and movement of data. Reliable precise knowledge about data has great potential improve reproducibility as well quality in biomedical research and, therefore, foster good scientific practice. However, despite increasing interest on technologies literature their implementation other disciplines, these have not yet been widely adopted research. </sec> <title>OBJECTIVE</title> The aim this scoping review was...