What AI Can Do for Your Country

Curated on Posted on by Stefaan Verhulst

Article by Jylana L. Sheats: “..Although most discussions of artificial intelligence focus on its impacts on business and research, AI is also poised to transform government in the United States and beyond. AI-guided disaster response is just one piece of the picture. The U.S. Department of Health and Human Services has an experimental AI program to diagnose COVID-19 and flu cases by analyzing the sound of patients coughing into their smartphones. The Department of Justice uses AI algorithms to help prioritize which tips in the FBI’s Threat Intake Processing System to act on first. Other proposals, still at the concept stage, aim to extend the applications of AI to improve the efficiency and effectiveness of nearly every aspect of public services. 

The early applications illustrate the potential for AI to make government operations more effective and responsive. They illustrate the looming challenges, too. The federal government will have to recruit, train, and retain skilled workers capable of managing the new technology, competing with the private sector for top talent. The government also faces a daunting task ensuring the ethical and equitable use of AI. Relying on algorithms to direct disaster relief or to flag high-priority crimes raises immediate concerns: What if biases built into the AI overlook some of the groups that most need assistance, or unfairly target certain populations? As AI becomes embedded into more government operations, the opportunities for misuse and unintended consequences will only expand…(More)”.

Use of large language models as a scalable approach to understanding public health discourse

Curated on Posted on by Stefaan Verhulst

Paper by Laura Espinosa and Marcel Salathé: “Online public health discourse is becoming more and more important in shaping public health dynamics. Large Language Models (LLMs) offer a scalable solution for analysing the vast amounts of unstructured text found on online platforms. Here, we explore the effectiveness of Large Language Models (LLMs), including GPT models and open-source alternatives, for extracting public stances towards vaccination from social media posts. Using an expert-annotated dataset of social media posts related to vaccination, we applied various LLMs and a rule-based sentiment analysis tool to classify the stance towards vaccination. We assessed the accuracy of these methods through comparisons with expert annotations and annotations obtained through crowdsourcing. Our results demonstrate that few-shot prompting of best-in-class LLMs are the best performing methods, and that all alternatives have significant risks of substantial misclassification. The study highlights the potential of LLMs as a scalable tool for public health professionals to quickly gauge public opinion on health policies and interventions, offering an efficient alternative to traditional data analysis methods. With the continuous advancement in LLM development, the integration of these models into public health surveillance systems could substantially improve our ability to monitor and respond to changing public health attitudes…(More)”.

Asserting the public interest in health data: On the ethics of data governance for biobanks and insurers

Curated on Posted on by Stefaan Verhulst

Paper by Kathryne Metcalf and Jathan Sadowski : “Recent reporting has revealed that the UK Biobank (UKB)—a large, publicly-funded research database containing highly-sensitive health records of over half a million participants—has shared its data with private insurance companies seeking to develop actuarial AI systems for analyzing risk and predicting health. While news reports have characterized this as a significant breach of public trust, the UKB contends that insurance research is “in the public interest,” and that all research participants are adequately protected from the possibility of insurance discrimination via data de-identification. Here, we contest both of these claims. Insurers use population data to identify novel categories of risk, which become fodder in the production of black-boxed actuarial algorithms. The deployment of these algorithms, as we argue, has the potential to increase inequality in health and decrease access to insurance. Importantly, these types of harms are not limited just to UKB participants: instead, they are likely to proliferate unevenly across various populations within global insurance markets via practices of profiling and sorting based on the synthesis of multiple data sources, alongside advances in data analysis capabilities, over space/time. This necessitates a significantly expanded understanding of the publics who must be involved in biobank governance and data-sharing decisions involving insurers…(More)”.

As AI-powered health care expands, experts warn of biases

Curated on Posted on by Stefaan Verhulst

Article by Marta Biino: “Google’s DeepMind artificial intelligence research laboratory and German pharma company BioNTech are both building AI-powered lab assistants to help scientists conduct experiments and perform tasks, the Financial Times reported.

It’s the latest example of how developments in artificial intelligence are revolutionizing a number of fields, including medicine. While AI has long been used in radiology, for image analysis, or oncology to classify skin lesions for example, as the technology continues to advance its applications are growing.

OpenAI’s GPT models have outperformed humans in making cancer diagnoses based on MRI reports and beat PhD-holders in standardized science tests, to name a few.

However, as AI’s use in health care expands, some fear the notoriously biased technology could carry negative repercussions for patients…(More)”.

Harnessing the feed: social media for mental health information and support 

Curated on Posted on by Stefaan Verhulst

Report by ReachOut: “…highlights how a social media ban could cut young people off from vital mental health support, including finding that 73 per cent of young people in Australia turn to social media when it comes to support for their mental health.

Based on research with over 2000 young people, the report found a range of benefits for young people seeking mental health support via social media (predominantly TikTok, YouTube and Instagram). 66 per cent of young people surveyed reported increased awareness about their mental health because of relevant content they accessed via social media, 47 per said they had looked for information about how to get professional mental health support on social media and 40 per cent said they sought professional support after viewing mental health information on social media. 

Importantly, half of young people with a probable mental health condition said that they were searching for mental health information or support on social media because they don’t have access to professional support. 

However, young people also highlighted a range of concerns about social media via the research. 38 per cent were deeply concerned about harmful mental health content they have come across on platforms and 43 per cent of the young people who sought support online were deeply concerned about the addictive nature of social media.  

The report highlights young people’s calls for social media to be safer. They want: an end to addictive features like infinite scroll, more control over the content they see, better labelling of mental health information from credible sources, better education and more mental health information provided across platforms…(More)”.

Harnessing digital footprint data for population health: a discussion on collaboration, challenges and opportunities in the UK

Curated on Posted on by Stefaan Verhulst

Paper by Romana Burgess et al: “Digital footprint data are inspiring a new era in population health and well-being research. Linking these novel data with other datasets is critical for future research wishing to use these data for the public good. In order to succeed, successful collaboration among industry, academics and policy-makers is vital. Therefore, we discuss the benefits and obstacles for these stakeholder groups in using digital footprint data for research in the UK. We advocate for policy-makers’ inclusion in research efforts, stress the exceptional potential of digital footprint research to impact policy-making and explore the role of industry as data providers, with a focus on shared value, commercial sensitivity, resource requirements and streamlined processes. We underscore the importance of multidisciplinary approaches, consumer trust and ethical considerations in navigating methodological challenges and further call for increased public engagement to enhance societal acceptability. Finally, we discuss how to overcome methodological challenges, such as reproducibility and sharing of learnings, in future collaborations. By adopting a multiperspective approach to outlining the challenges of working with digital footprint data, our contribution helps to ensure that future research can navigate these challenges effectively while remaining reproducible, ethical and impactful…(More)”

Climate and health data website launched

Curated on Posted on by Stefaan Verhulst

Article by Susan Cosier: “A new website of data resources, tools, and training materials that can aid researchers in studying the consequences of climate change on the health of communities nationwide is now available. At the end of July, NIEHS launched the Climate and Health Outcomes Research Data Systems (CHORDS) website, which includes a catalog of environmental and health outcomes data from various government and nongovernmental agencies.

The website provides a few resources of interest, including a catalog of data resources to aid researchers in finding relevant data for their specific research projects; an online training toolkit that provides tutorials and walk-throughs of downloading, integrating, and visualizing health and environmental data; a listing of publications of note on wildfire and health research; and links to existing resources, such as the NIEHS climate change and health glossary and literature portal.

The catalog includes a listing of dozens of data resources provided by different federal and state environmental and health sources. Users can sort the listing based on environmental and health measures of interest — such as specific air pollutants or chemicals — from data providers including NASA and the U.S. Environmental Protection Agency with many more to come…(More)”.

Utilizing big data without domain knowledge impacts public health decision-making

Curated on Posted on by Stefaan Verhulst

Paper by Miao Zhang, Salman Rahman, Vishwali Mhasawade and Rumi Chunara: “…New data sources and AI methods for extracting information are increasingly abundant and relevant to decision-making across societal applications. A notable example is street view imagery, available in over 100 countries, and purported to inform built environment interventions (e.g., adding sidewalks) for community health outcomes. However, biases can arise when decision-making does not account for data robustness or relies on spurious correlations. To investigate this risk, we analyzed 2.02 million Google Street View (GSV) images alongside health, demographic, and socioeconomic data from New York City. Findings demonstrate robustness challenges; built environment characteristics inferred from GSV labels at the intracity level often do not align with ground truth. Moreover, as average individual-level behavior of physical inactivity significantly mediates the impact of built environment features by census tract, intervention on features measured by GSV would be misestimated without proper model specification and consideration of this mediation mechanism. Using a causal framework accounting for these mediators, we determined that intervening by improving 10% of samples in the two lowest tertiles of physical inactivity would lead to a 4.17 (95% CI 3.84–4.55) or 17.2 (95% CI 14.4–21.3) times greater decrease in the prevalence of obesity or diabetes, respectively, compared to the same proportional intervention on the number of crosswalks by census tract. This study highlights critical issues of robustness and model specification in using emergent data sources, showing the data may not measure what is intended, and ignoring mediators can result in biased intervention effect estimates…(More)”

Collaboration in Healthcare: Implications of Data Sharing for Secondary Use in the European Union

Curated on Posted on by Stefaan Verhulst

Paper by Fanni Kertesz: “The European healthcare sector is transforming toward patient-centred and value-based healthcare delivery. The European Health Data Space (EHDS) Regulation aims to unlock the potential of health data by establishing a single market for its primary and secondary use. This paper examines the legal challenges associated with the secondary use of health data within the EHDS and offers recommendations for improvement. Key issues include the compatibility between the EHDS and the General Data Protection Regulation (GDPR), barriers to cross-border data sharing, and intellectual property concerns. Resolving these challenges is essential for realising the full potential of health data and advancing healthcare research and innovation within the EU…(More)”.

Using internet search data as part of medical research

Curated on Posted on by Stefaan Verhulst

Blog by Susan Thomas and Matthew Thompson: “…In the UK, almost 50 million health-related searches are made using Google per year. Globally there are 100s of millions of health-related searches every day. And, of course, people are doing these searches in real-time, looking for answers to their concerns in the moment. It’s also possible that, even if people aren’t noticing and searching about changes to their health, their behaviour is changing. Maybe they are searching more at night because they are having difficulty sleeping or maybe they are spending more (or less) time online. Maybe an individual’s search history could actually be really useful for researchers. This realisation has led medical researchers to start to explore whether individuals’ online search activity could help provide those subtle, almost unnoticeable signals that point to the beginning of a serious illness.

Our recent review found 23 studies have been published so far that have done exactly this. These studies suggest that online search activity among people later diagnosed with a variety of conditions ranging from pancreatic cancer and stroke to mood disorders, was different to people who did not have one of these conditions.

One of these studies was published by researchers at Imperial College London, who used online search activity to identify signals of women with gynaecological malignancies. They found that women with malignant (e.g. ovarian cancer) and benign conditions had different search patterns, up to two months prior to a GP referral. 

Pause for a moment, and think about what this could mean. Ovarian cancer is one of the most devastating cancers women get. It’s desperately hard to detect early – and yet there are signals of this cancer visible in women’s internet searches months before diagnosis?…(More)”.