From Pixels to Principles: A Decade of Progress and Landscape in Trustworthy Computer Vision.

The rapid development of computer vision technologies and applications has brought forth a range of social and ethical challenges. Due to the unique characteristics of visual technology in terms of data modalities and application scenarios, computer vision poses specific ethical issues. However, the majority of existing literature either addresses artificial intelligence as a whole or pays particular attention to natural language processing, leaving a gap in specialized research on ethical issues and systematic solutions in the field of computer vision. This paper utilizes bibliometrics and text-mining techniques to quantitatively analyze papers from prominent academic conferences in computer vision over the past decade. It first reveals the developing trends and specific distribution of attention regarding trustworthy aspects in the computer vision field, as well as the inherent connections between ethical dimensions and different stages of visual model development. A life-cycle framework regarding trustworthy computer vision is then presented by making the relevant trustworthy issues, the operation pipeline of AI models, and viable technical solutions interconnected, providing researchers and policymakers with references and guidance for achieving trustworthy CV. Finally, it discusses particular motivations for conducting trustworthy practices and underscores the consistency and ambivalence among various trustworthy principles and technical attributes.

Ageism in COVID-Related Newspaper Coverage: The First Month of a Pandemic.

OBJECTIVES: Media sources have consistently described older adults as a medically vulnerable population during the coronavirus disease 2019 (COVID-19) pandemic, yet a lack of concern over their health and safety has resulted in dismissal and devaluation. This unprecedented situation highlights ongoing societal ageism and its manifestations in public discourse. This analysis asks how national news sources performed explicit and implicit ageism during the first month of the pandemic. METHOD: Using content and critical discourse analysis methods, we analyzed 287 articles concerning older adults and COVID-19 published between March 11 and April 10, 2020, in 4 major U.S.-based newspapers. RESULTS: Findings indicate that while ageism was rarely discussed explicitly, ageist bias was evident in implicit reporting patterns (e.g., frequent use of the term "elderly," portrayals of older adults as "vulnerable"). Infection and death rates and institutionalized care were among the most commonly reported topics, providing a limited portrait of aging during the pandemic. The older "survivor" narrative offers a positive alternative by suggesting exceptional examples of resilience and grit. However, the survivor narrative may also implicitly place blame on those unable to survive or thrive in later life. DISCUSSION: This study provides insight for policy makers, researchers, and practitioners exploring societal perceptions of older adults and how these perceptions are disseminated and maintained by the media.

Paths to social licence for tracking-data analytics in university research and services.

While tracking-data analytics can be a goldmine for institutions and companies, the inherent privacy concerns also form a legal, ethical and social minefield. We present a study that seeks to understand the extent and circumstances under which tracking-data analytics is undertaken with social licence-that is, with broad community acceptance beyond formal compliance with legal requirements. Taking a University campus environment as a case, we enquire about the social licence for Wi-Fi-based tracking-data analytics. Staff and student participants answered a questionnaire presenting hypothetical scenarios involving Wi-Fi tracking for university research and services. Our results present a Bayesian logistic mixed-effects regression of acceptability judgements as a function of participant ratings on 11 privacy dimensions. Results show widespread acceptance of tracking-data analytics on campus and suggest that trust, individual benefit, data sensitivity, risk of harm and institutional respect for privacy are the most predictive factors determining this acceptance judgement.

First do no harm: An exploration of researchers' ethics of conduct in Big Data behavioral studies.

Research ethics has traditionally been guided by well-established documents such as the Belmont Report and the Declaration of Helsinki. At the same time, the introduction of Big Data methods, that is having a great impact in behavioral research, is raising complex ethical issues that make protection of research participants an increasingly difficult challenge. By conducting 39 semi-structured interviews with academic scholars in both Switzerland and United States, our research aims at exploring the code of ethics and research practices of academic scholars involved in Big Data studies in the fields of psychology and sociology to understand if the principles set by the Belmont Report are still considered relevant in Big Data research. Our study shows how scholars generally find traditional principles to be a suitable guide to perform ethical data research but, at the same time, they recognized and elaborated on the challenges embedded in their practical application. In addition, due to the growing introduction of new actors in scholarly research, such as data holders and owners, it was also questioned whether responsibility to protect research participants should fall solely on investigators. In order to appropriately address ethics issues in Big Data research projects, education in ethics, exchange and dialogue between research teams and scholars from different disciplines should be enhanced. In addition, models of consultancy and shared responsibility between investigators, data owners and review boards should be implemented in order to ensure better protection of research participants.

Privacy-Preserving Process Mining in Healthcare.

Process mining has been successfully applied in the healthcare domain and has helped touncover various insights for improving healthcare processes. While the benefits of process miningare widely acknowledged, many people rightfully have concerns about irresponsible uses of personaldata. Healthcare information systems contain highly sensitive information and healthcare regulationsoften require protection of data privacy. The need to comply with strict privacy requirements mayresult in a decreased data utility for analysis. Until recently, data privacy issues did not get muchattention in the process mining community; however, several privacy-preserving data transformationtechniques have been proposed in the data mining community. Many similarities between datamining and process mining exist, but there are key differences that make privacy-preserving datamining techniques unsuitable to anonymise process data (without adaptations). In this article, weanalyse data privacy and utility requirements for healthcare process data and assess the suitabilityof privacy-preserving data transformation methods to anonymise healthcare data. We demonstratehow some of these anonymisation methods affect various process mining results using three publiclyavailable healthcare event logs. We describe a framework for privacy-preserving process mining thatcan support healthcare process mining analyses. We also advocate the recording of privacy metadatato capture information about privacy-preserving transformations performed on an event log.

Systematic Review of Privacy-Preserving Distributed Machine Learning From Federated Databases in Health Care.

Big data for health care is one of the potential solutions to deal with the numerous challenges of health care, such as rising cost, aging population, precision medicine, universal health coverage, and the increase of noncommunicable diseases. However, data centralization for big data raises privacy and regulatory concerns.Covered topics include (1) an introduction to privacy of patient data and distributed learning as a potential solution to preserving these data, a description of the legal context for patient data research, and a definition of machine/deep learning concepts; (2) a presentation of the adopted review protocol; (3) a presentation of the search results; and (4) a discussion of the findings, limitations of the review, and future perspectives.Distributed learning from federated databases makes data centralization unnecessary. Distributed algorithms iteratively analyze separate databases, essentially sharing research questions and answers between databases instead of sharing the data. In other words, one can learn from separate and isolated datasets without patient data ever leaving the individual clinical institutes.Distributed learning promises great potential to facilitate big data for medical application, in particular for international consortiums. Our purpose is to review the major implementations of distributed learning in health care.

Perspectives of Australian policy-makers on the potential benefits and risks of technologically enhanced communicable disease surveillance - a modified Delphi survey.

BACKGROUND: Event-based social media monitoring and pathogen whole genome sequencing (WGS) will enhance communicable disease surveillance research and systems. If linked electronically and scanned systematically, the information provided by these technologies could be mined to uncover new epidemiological patterns and associations much faster than traditional public health approaches. The benefits of earlier outbreak detection are significant, but implementation could be opposed in the absence of a social licence or if ethical and legal concerns are not addressed. METHODS: A three-phase mixed-method Delphi survey with Australian policy-makers, health practitioners and lawyers (n = 44) was conducted to explore areas of consensus and disagreement over (1) key policy and practical issues raised by the introduction of novel communicable disease surveillance programmes; and (2) the most significant and likely risks from using social media content and WGS technologies in epidemiological research and outbreak investigations. RESULTS: Panellists agreed that the integration of social media monitoring and WGS technologies into communicable disease surveillance systems raised significant issues, including impacts on personal privacy, medicolegal risks and the potential for unintended consequences. Notably, their concerns focused on how these technologies should be used, rather than how the data was collected. Panellists held that social media users should expect their posts to be monitored in the interests of public health, but using those platforms to contact identified individuals was controversial. The conditions of appropriate use of pathogen WGS in epidemiological research and investigations was also contentious. Key differences amongst participants included the necessity for consent before testing and data-linkage, thresholds for action, and the legal and ethical importance of harms to individuals and commercial entities. The erosion of public trust was seen as the most significant risk from the systematic use of these technologies. CONCLUSIONS: Enhancing communicable disease surveillance with social-media monitoring and pathogen WGS may cause controversy. The challenge is to determine and then codify how these technologies should be used such that the balance between individual risk and community benefit is widely accepted. Participants agreed that clear guidelines for appropriate use that address legal and ethical concerns need to be developed in consultation with relevant experts and the broader Australian public.

Engineering Bias in AI.

After working at Apple designing circuits and signal processing algorithms for products including the first iPad, Timnit Gebru (Figure 1) received her Ph.D. from the Stanford Artificial Intelligence Laboratory in the area of computer vision. She recently completed a postdoc with Microsoft Research in the FATE (Fairness, Transparency, Accountability, and Ethics in Artificial Intelligence (AI)) group, was a cofounder of Black in AI, and is currently working as a research scientist in the Ethical AI team at Google. Her research in algorithmic bias and the ethical implications of data mining have appeared in multiple publications, including The New York Times and The Economist. IEEE Pulse recently spoke with Gebru about the role societal bias plays in engineering AI, the deficits and dangers in the field caused by limited diversity, and the challenges inherent in addressing these complex issues.

Toward an accelerated adoption of data-driven findings in medicine : Research, skepticism, and the need to speed up public visibility of data-driven findings.

To accelerate the adoption of a new method with a high potential to replace or extend an existing, presumably less accurate, medical scoring system, evaluation should begin days after the new concept is presented publicly, not years or even decades later. Metaphorically speaking, as chameleons capable of quickly changing colors to help their bodies adjust to changes in temperature or light, health-care decision makers should be capable of more quickly evaluating new data-driven insights and tools and should integrate the highest performing ones into national and international care systems. Doing so is essential, because it will truly save the lives of many individuals.

Documento de información sobre la utilización de los datos de la historia clínica con fines de investigación biomédica en contextos de bigdata / Information sheet on the use of clinical record for biomedical research purposes in bigdata contexts

No disponible

Nuevos desafíos en torno al big data / New challenges around big data

Los nuevos reglamentos europeos sobre ensayos clínicos, dispositivos médicos y el nuevo Reglamento Europeo sobre protección de datos, incorporan varios preceptos para garantizar el derecho a la vida privada y la protección de datos en materia de salud. Sin embargo, la fragmentación de la regulación, el riesgo de sufrir ciber-ataques y violaciones de seguridad, las filtraciones masivas de big data, o el uso no autorizado de datos biométricos nos llevan a poner en duda el papel predominante que la regulación otorga al consentimiento previo del propietario en la cesión de los datos personales como clave del sistema. En este sentido, las normas de protección de datos generales del nuevo reglamento prohíben el tratamiento de los datos personales relativos a la salud, pero las numerosas excepciones a esta regla general pueden limitar los derechos del interesado. Además, la falta de medidas técnicas y organizativas comunes para garantizar el respeto del principio de minimización de datos y la falta de obligación de aplicar medidas de compatibilidad para intercambiar, cuando sea necesario, los datos obtenidos en los Estados miembros pueden poner en grave riesgo los beneficios de la regulación y amenazar la efectividad del derecho a la privacidad

New European Regulations on clinical trials, medical devices or the European Regulation on data protection, Regulation (EU) 2016/679 of the European Parliament and of the Council of April 27, 2016, incorporate several rules to ensure the right to privacy and the data protection in the field of health. However, the fragmentation of the regulation, the risk of cyber-attacks and security breaches, the massive leaks of big data or the unauthorized use of biometric data, lead us to question about the dominant role that the regulation gives to the prior consent of the owner in the transfer of personal data as a key of the system. In this sense, the rules of protection of general data of the new Regulation prohibit the treatment of the personal data relating to health, but numerous exceptions to this general rule, may limit the rights of the person concerned. In addition, the lack of common technical and organizational measures to ensure the respect of the principle of data minimization and lack of obligation to implement measures of support for exchange, where necessary, the data obtained in the States Members can put at risk the benefits of regulation and threaten the effectiveness of the right to privacy

Utilización de tecnología Big Data en investigación clínica / Use of Big Data technology in clinical research

La utilización de la tecnología Big data en la práctica clínica puede suponer grandes avances para el sistema sanitario, tanto para la asistencia como para la investigación clínica. Para conseguir este objetivo es imprescindible la integración de las diferentes fuentes de información actualmente disponibles. De forma simultánea a estas ventajas potenciales, la aplicación de estas tecnologías puede suponer una fuerte amenaza para la intimidad, por lo que es imprescindible disponer de medidas adecuadas de control de la información, así como implantar procedimientos adecuados, transparentes y seguros, que garanticen el máximo nivel de confidencialidad asegurando el respeto a los derechos y libertades de las personas. Para poder utilizar de forma segura esta tecnología es imprescindible disponer de un marco ético y jurídico adecuado, que permita compaginar la protección de datos con la investigación clínica relevante y la mejora de la asistencia

The use of Big data technology in clinical practice can mean great advances for the healthcare system, both for care and for clinical research. To achieve this goal, it is essential to integrate the different sources of information currently available. Simultaneously with these potential advantages, the application of these technologies can pose a strong threat to privacy, so it is essential to have adequate control measures for information, as well as to implement adequate, transparent and secure procedures, which guarantee the highest level of confidentiality ensuring respect for the rights and freedoms of individuals. In order to be able to use this technology safely, it is essential to have an adequate ethical and legal framework that allows data protection to be combined with relevant clinical research and improved care

Revisión de las categorías jurídicas de la normativa europea ante la tecnología del big data aplicada a la salud / Revision of the legal categories of the European regulation in the face of big data technology applied to health

El nuevo RGPD dedica una mayor atención específica a los datos personales relativos a la salud, lo cual era estrictamente necesario. Además, se incluyen de forma explícita y por primera vez varias referencias a los datos genéticos en cuanto datos relativos a la salud, aunque separados de éstos. La posición actual de la UE y de los EM sobre el estatuto jurídico de los datos relativos a la salud ha cambiado sensiblemente, pues, aunque éstos conservan su condición de datos "sensibles", esto es, de datos que gozan de una protección jurídica especial, se ha decidido también facilitar el acceso a estos datos por parte de los diversos profesionales de la salud que tengan que prestar su actividad asistencial con el fin de ganar en eficacia y en rapidez respecto a dicho acceso. Mientras que en este supuesto se han querido primar los intereses del propio titular de los datos en relación con su salud, en el caso de la investigación relativa a la salud o biomédica con la eliminación o relajación de ciertos requisitos se ha dado preferencia al interés social que representa la misma frente al derecho individual a la protección de los datos personales, en la medida en que los resultados y avances científicos en el sector de la salud contribuyen al bienestar de la colectividad. Es obvio que otras disposiciones de carácter más general, que atienden también a situaciones nuevas o cambiantes, y por tanto a las necesidades jurídicas actuales, serán aplicables asimismo a los datos relativos a la salud; así, respecto al tratamiento masivo de datos y el flujo transnacional de datos, que han experimentado modificaciones relevantes con el nuevo marco legal europeo y, como es sabido, ambos supuestos son de extraordinario interés para los datos relativos a la salud. Coherentemente, la legislación interna de nuestro país sobre protección de datos personales ha sido objeto de revisión y de adaptación parlamentarias al RGPD mediante la promulgación de una nueva ley orgánica. En este artículo el autor estudia algunos conceptos y categorías jurídicas nuevos o revisados por la nueva regulación europea o que requieren un enfoque diferente, con el fin de delimitar su verdadero significado y alcance en la actualidad. Para este fin, tiene presente la nueva regulación estatal sobre protección de datos cuando resulta pertinente

The new GDPR devotes more specific attention to personal data relating to health, which was strictly necessary. In addition, for the first time a number of references to genetic data are explicitly included as health-related data but separate from them. The current position of the EU and the MS on the legal status of health data has changed significantly, even though they retain their status as "sensitive" data, i.e. data enjoying special legal protection, it has also been decided to facilitate access to these data by the various health professionals who have to provide care in order to increase the efficiency and speed of such access. While in this case the interests of the data subject in relation to his/her health have been prioritised, in the case of health or biomedical research with the elimination or relaxation of certain requirements, preference has been given to the social interest which it represents over the individual right to the protection of personal data, insofar as scientific results and advances in the health sector contribute to the well-being of the community. It is obvious that other provisions of a more general nature, which also deal with new or changing situations, and therefore with current legal needs, will also apply to data relating to health; thus, with regard to the massive processing of data and the transnational flow of data, which have undergone relevant modifications with the new European legal framework and, as is known, both assumptions are of extraordinary interest for data relating to health. Consistently, our country's internal legislation on the protection of personal data has been subject to parliamentary revision and adaptation to the GDPR through the enactment of a new fundamental law. In this paper the author studies some legal concepts and categories that are new or revised by the new European regulation or that require a different approach, in order to delimit their true meaning and scope at present. To this end, the author takes into account the new state regulation on data protection when it is relevant for that purpose

Machine learning in medicine: Addressing ethical challenges.

Effy Vayena and colleagues argue that machine learning in medicine must offer data protection, algorithmic transparency, and accountability to earn the trust of patients and clinicians.

[Chapter 1. From the study of risks to the translation of the ethical issues of Big Data in Health]. / Chapitre 1. De l?étude des risques à la traduction des enjeux éthiques des Big Data en santé.

Big Data substantially disrupt the medical microcosm up to challenge the paradigms of Medicine Hippocrates as we previously know. Therefore, a reflection on the study of risks associated with ethical issues around personal health data is imposed on us. Our study is based on many field surveys, interviews targeted at different actors, as well as a literature search on the subject. This work led to the realization of an innovative method of alignment of concepts of ontology of the risks to those of ontology of ethical objectives requirements of Big Data in Health. The aim is to make sense and recommendations to realization, implementation and use of personal data in order to better control it.

[Chapter 7. Big Data or the illusion of a synthesis by aggregation. Epistemological, ethical and political critics]. / Chapitre 7. Big data ou l?illusion d?une synthèse par agrégation. Une critique épistémologique, éthique et politique.

In this article, we propose a critical approach to the big data phenomenon by deconstructing the methodological principle that structures its logic : the principle of aggregation. Our hypothesis is upstream of the critics who make the use of big data a new mode of government. Aggregation, as a mode of processing the heterogeneity of data, structures the thinking big data, it is its very logic. Fragmentation in order to better aggregate, to aggregate to better fragment, a dialectic based on a presumption of generalized aggregability and on the claim to make aggregation the preferred route for the production of new syntheses. We proceed in three steps to deconstruct this idea and undo the claim of aggregation to assert itself as a new way to produce knowledge, as a new synthesis of identity and finally as a new model of solidarity. Each time we show that these attempts at aggregation fail to produce their objects : no knowledge, no identity, no solidarity can result from a process of amalgamation. In all three cases, aggregation is always accompanied by a moment of fragmentation whose dissociation, dislocation and separation are different figures. The bet we are making then is to make hesitate what presents itself as a new way of thinking man and the world.

Electronic health records to facilitate clinical research.

Electronic health records (EHRs) provide opportunities to enhance patient care, embed performance measures in clinical practice, and facilitate clinical research. Concerns have been raised about the increasing recruitment challenges in trials, burdensome and obtrusive data collection, and uncertain generalizability of the results. Leveraging electronic health records to counterbalance these trends is an area of intense interest. The initial applications of electronic health records, as the primary data source is envisioned for observational studies, embedded pragmatic or post-marketing registry-based randomized studies, or comparative effectiveness studies. Advancing this approach to randomized clinical trials, electronic health records may potentially be used to assess study feasibility, to facilitate patient recruitment, and streamline data collection at baseline and follow-up. Ensuring data security and privacy, overcoming the challenges associated with linking diverse systems and maintaining infrastructure for repeat use of high quality data, are some of the challenges associated with using electronic health records in clinical research. Collaboration between academia, industry, regulatory bodies, policy makers, patients, and electronic health record vendors is critical for the greater use of electronic health records in clinical research. This manuscript identifies the key steps required to advance the role of electronic health records in cardiovascular clinical research.