Priorities for successful use of artificial intelligence by public health organizations: a literature review

BMC Public Health, Год журнала: 2022, Номер 22(1)

Опубликована: Ноя. 22, 2022

Artificial intelligence (AI) has the potential to improve public health's ability promote health of all people in communities. To successfully realize this and use AI for functions it is important organizations thoughtfully develop strategies implementation. Six key priorities successful technologies by are discussed: 1) Contemporary data governance; 2) Investment modernized analytic infrastructure procedures; 3) Addressing skills gap workforce; 4) Development strategic collaborative partnerships; 5) Use good practices transparency reproducibility, and; 6) Explicit consideration equity bias.

Язык: Английский

---

Emerging technologies in public health campaigns: Artificial intelligence and big data

International Journal of Science and Research Archive, Год журнала: 2024, Номер 11(1), С. 478 - 487

Опубликована: Янв. 26, 2024

This research explores the integration of Artificial Intelligence (AI) and Big Data into public health campaigns, envisioning a future where precision, personalization, proactive interventions redefine healthcare. Analyzing transformative potential challenges, study examines AI's role in disease surveillance, diagnostics, predictive modeling, alongside Data's contributions to personalized comprehensive understanding. Ethical considerations, digital divide, regulatory frameworks are central necessitating delicate balance between innovation responsibility. The conclusion foresees healthcare landscape AI enhance effectiveness promising characterized by equitable, data-driven, resilient approaches address emerging challenges.

Язык: Английский

---

Leveraging Administrative Health Databases to Address Health Challenges in Farming Populations: Scoping Review and Bibliometric Analysis (1975-2024)

JMIR Public Health and Surveillance, Год журнала: 2025, Номер 11, С. e62939 - e62939

Опубликована: Янв. 9, 2025

Background Although agricultural health has gained importance, to date, much of the existing research relies on traditional epidemiological approaches that often face limitations related sample size, geographic scope, temporal coverage, and range events examined. To address these challenges, a complementary approach involves leveraging reusing data beyond its original purpose. Administrative databases (AHDs) are increasingly reused in population-based digital public health, especially for populations such as farmers, who distinct environmental risks. Objective We aimed explore reuse AHDs addressing issues within farming by summarizing current landscape AHD-based identifying key areas interest, gaps, unmet needs. Methods conducted scoping review bibliometric analysis using PubMed Web Science. Building upon previous reviews research, we comprehensive literature search 72 terms population AHDs. identify hot spots, directions, used keyword frequency, co-occurrence, thematic mapping. also explored profile exposome mapping co-occurrences between factors outcomes. Results Between 1975 April 2024, 296 publications across 118 journals, predominantly from high-income countries, were identified. Nearly one-third associated with well-established cohorts, Agriculture Cancer Agricultural Health Study. The most frequently included disease registers (158/296, 53.4%), electronic records (124/296, 41.9%), insurance claims (106/296, 35.8%), (95/296, 32.1%), hospital discharge (41/296, 13.9%). Fifty (16.9%) studies involved >1 million participants. broad exposure proxies used, (254/296, 85.8%) relied proxies, which failed capture specifics tasks. Research remains underexplored, predominant focus specific external exposome, particularly pesticide exposure. A limited have been examined, primarily cancer, mortality, injuries. Conclusions increasing use holds major potential advance populations. However, substantial gaps persist, low-income regions among underrepresented subgroups, women, children, contingent workers. Emerging issues, including per- polyfluoroalkyl substances, biological agents, microbiome, microplastics, climate change, warrant further research. Major persist understanding various conditions, cardiovascular, reproductive, ocular, sleep-related, age-related, autoimmune diseases. Addressing overlooked is essential comprehending risks faced communities guiding policies. Within this context, promoting conjunction other sources (eg, mobile social data, wearables) artificial intelligence approaches, represents promising avenue future exploration.

Язык: Английский

---

Integrating Ethics within Machine Learning Courses

ACM Transactions on Computing Education, Год журнала: 2019, Номер 19(4), С. 1 - 26

Опубликована: Авг. 2, 2019

This article establishes and addresses opportunities for ethics integration into Machine-learning (ML) courses. Following a survey of the history computing current need ethical consideration within ML, we consider state ML education via an exploratory analysis course syllabi in programs. The results reveal that though is part overall educational landscape these programs, it not frequently core technical To help address this gap, offer preliminary framework, developed systematic literature review, relevant questions should be addressed project. A pilot study with 85 students confirms framework helped them identify articulate key considerations their projects. Building from work, also provide three example modules bring thinking directly learning content. Collectively, research demonstrates: (1) to taught as integrated coursework, (2) structured set useful identifying addressing potential issues project, (3) novel models examples how practically teach without sacrificing An additional by-product collection recent publications emerging field education.

Язык: Английский

---

Disruptive Technologies for Environment and Health Research: An Overview of Artificial Intelligence, Blockchain, and Internet of Things

International Journal of Environmental Research and Public Health, Год журнала: 2019, Номер 16(20), С. 3847 - 3847

Опубликована: Окт. 11, 2019

The purpose of this descriptive research paper is to initiate discussions on the use innovative technologies and their potential support development pan-Canadian monitoring surveillance activities associated with environmental impacts health within system. Its primary aim provide a review disruptive current uses in environment healthcare. Drawing extensive experience population-level through technology, knowledge from prior projects field, conducting technologies, meant serve as initial steps toward better understanding area. In doing so, we hope be able assess which might best leveraged advance unique intersection environment. This first outlines at public environment, particular, Artificial Intelligence (AI), Blockchain, Internet Things (IoT). provides description for each these along summary applications, challenges one face adopting them. Thereafter, high-level reference architecture, that addresses described could potentially incorporated into system, conceived presented.

Язык: Английский

---

Values, challenges and future directions of big data analytics in healthcare: A systematic review

Social Science & Medicine, Год журнала: 2019, Номер 241, С. 112533 - 112533

Опубликована: Сен. 10, 2019

Язык: Английский

---

Artificial Intelligence in Health Care: Current Applications and Issues

Journal of Korean Medical Science, Год журнала: 2020, Номер 35(42)

Опубликована: Янв. 1, 2020

In recent years, artificial intelligence (AI) technologies have greatly advanced and become a reality in many areas of our daily lives.In the health care field, numerous efforts are being made to implement AI technology for practical medical treatments.With rapid developments machine learning algorithms improvements hardware performances, is expected play an important role effectively analyzing utilizing extensive amounts data.However, has various unique characteristics that different from existing technologies.Subsequently, there number need be supplemented within current system utilized more frequently care.In addition, practitioners public accept still low; moreover, concerns regarding safety reliability

Язык: Английский

---

Reflection on modern methods: when worlds collide—prediction, machine learning and causal inference

International Journal of Epidemiology, Год журнала: 2019, Номер 49(6), С. 2058 - 2064

Опубликована: Июнь 11, 2019

Abstract Causal inference requires theory and prior knowledge to structure analyses, is not usually thought of as an arena for the application prediction modelling. However, contemporary causal methods, premised on counterfactual or potential outcomes approaches, often include processing steps before final estimation step. The purposes this paper are: (i) overview recent emergence underpinning in methods a useful perspective (ii) explore role machine learning (as one approach ‘best prediction’) inference. covered propensity scores, inverse probability treatment weights (IPTWs), G computation targeted maximum likelihood (TMLE). Machine has been used more scores TMLE, there increased use IPTWs.

Язык: Английский

---

Making Messy Data Work for Conservation

One Earth, Год журнала: 2020, Номер 2(5), С. 455 - 465

Опубликована: Май 1, 2020

Conservationists increasingly use unstructured observational data, such as citizen science records or ranger patrol observations, to guide decision making. These datasets are often large and relatively cheap collect, they have enormous potential. However, the resulting data generally "messy," their can incur considerable costs, some of which hidden. We present an overview opportunities limitations associated with messy by explaining how preferences, skills, incentives collectors affect quality information contain investment required unlock Drawing widely from across sciences, we break down elements observation process in order highlight likely sources bias error while emphasizing importance cross-disciplinary collaboration. propose a framework for appraising those engaging these types dataset make them work conservation broader sustainability applications. The world's ecosystems face daunting array threats, including habitat loss, overexploitation, invasive species, pollution, climate change.1Doherty T.S. Dickman C.R. Nimmo D.G. Ritchie E.G. Multiple multiplying threats? Interactions between predators other ecological disturbances.Biol. Conserv. 2015; 190: 60-68Crossref Scopus (132) Google Scholar, 2Young H.S. McCauley D.J. Galetti M. Dirzo R. Patterns, causes, consequences anthropocene defaunation.Annu. Rev. Ecol. Evol. Syst. 2016; 47: 333-358Crossref (160) 3Venter O. Sanderson E.W. Magrach A. Allan J.R. Beher J. Jones K.R. Possingham H.P. Laurance W.F. Wood P. Fekete B.M. et al.Sixteen years change global terrestrial human footprint implications biodiversity conservation.Nat. Commun. 7: 12558Crossref PubMed (550) 4Benítez-López Alkemade Schipper Ingram D. Verweij Eikelboom Huijbregts impact hunting on tropical mammal bird populations.Science. 2017; 356: 180-183Crossref (186) Scholar Robust must be cornerstone scientists all stripes seeking understand dynamics environmental map out pathways toward sustainability.5Cadotte M.W. Barlow Nuñez M.A. Pettorelli N. Stephens P.A. Solving problems Anthropocene: need bring novel theoretical advances into applied ecology fold.J. Appl. 54: 1-6Crossref (22) Practical decisions promotion health evidence based, interventions no exception,6Sutherland W.J. Pullin A.S. Dolman P.M. Knight T.M. evidence-based conservation.Trends 2004; 19: 305-308Abstract Full Text PDF (1121) 7Walsh J.C. Dicks L.V. Sutherland effect scientific practitioners' management decisions.Conserv. Biol. 29: 88-98Crossref (102) 8Milner-Gulland E.J. Cadotte Hulme P.E. Kerby G. Whittingham M.J. Ensuring has impact.J. 2012; 49: 1-5Crossref (23) but gathering that via primary collection within formal study design is expensive, time consuming, impractical.9Gardner T.A. Araujo I.S. Ávila-Pires T.C. Bonaldo A.B. Costa J.E. Esposito M.C. Ferreira Hawes Hernandez M.I. cost-effectiveness surveys forests.Ecol. Lett. 2008; 11: 139-150Crossref (386) Scholar,10Vijapure T. Sukumaran S. Optimization taxonomic resolution indicator taxon cost-effective monitoring: perspectives heterogeneous coastline.J. Environ. Manage. 2019; 247: 474-483Crossref (4) Confronted complex restrictive budgets, governments conservationists draw rapidly growing body semi-structured monitoring trends assessing interventions.11Follett Strezov V. An analysis based research: usage publication patterns.PLoS One. 10: e0143687Crossref (153) 12Johnston Fink Reynolds M.D. Hochachka W.M. Sullivan B.L. Bruns N.E. Hallstein E. Merrifield M.S. Matsumoto Kelling Abundance models improve spatial temporal prioritization resources.Ecol. 25: 1749-1756Crossref (78) 13Woodcock B.A. Isaac N.J.B. Bullock J.M. Roy D.B. Garthwaite Crowe Pywell R.F. Impacts neonicotinoid long-term population changes wild bees England.Nat. 12459Crossref (206) high-volume, been subject number recent reviews data-generation potential social media online technologies,14Muller C.L. Chapman L. Johnston Kidd C. Illingworth Foody Overeem Leigh R.R. Crowdsourcing atmospheric sciences: current status future potential.Int. Climatol. 35: 3185-3203Crossref (167) phenomenon big data,15Mooney S.J. Pejaver Big public health: terminology, machine learning, privacy.Annu. Public Health. 2018; 39: 95-112Crossref (83) Scholar,16Hochachka Hutchinson R.A. Sheldon Wong W.K. Data-intensive broad-scale science.Trends 27: 130-137Abstract (231) understanding of, participation in, science.14Muller 15Mooney 16Hochachka 17Bonney Phillips T.B. Ballard H.L. Enck J.W. Can enhance science?.Public Underst. Sci. 2-16Crossref (256) 18Ballard Dixon C.G.H. Harris E.M. Youth-focused science: examining role learning agency conservation.Biol. 208: 65-75Crossref limited attention paid mechanisms arise ways issues may anticipated (bias avoidance) overcome mitigation). Here, umbrella term "messy data" describe whose does not conform thus potentially unmeasured (Box 1). They typically generated processes designed either (1) separate purpose, wherein secondary (e.g., patrols), (2) generating where and/or opportunistic many projects). "observers" cover gatherers any form whom unwitting, unpaid, collecting adjunct objective. Within this definition "messy" exists wide range (Figure For example, projects Cornell Laboratory Ornithology's eBird, survey designers lack control over behavior observers sufficient resources enough metadata, sophisticated statistical modeling account aspects bias.20Sullivan Aycrigg J.L. Barry J.H. Bonney R.E. Cooper C.B. Damoulas Dhondt A.A. Dietterich Farnsworth al.The eBird enterprise: integrated approach development application science.Biol. 2014; 169: 31-40Crossref (384) Other datasets, contrast, about (or generators), producing biases much harder tackle. This latter group includes herbaria museums, patrols, illegal wildlife trade seizures at international borders, crowd-sensing posts.21Moore J.F. Mulindahabi F. Masozera M.K. Nichols J.D. Hines Turikunkiko Oli Hayward Are patrols effective reducing poaching-related threats protected areas?.J. 55: 99-107Crossref (30) 22Hinsley Lee T.E. Harrison Roberts D.L. Estimating extent structure horticultural orchids media.Conserv. 30: 1038-1047Crossref (52) 23Runhovde S.R. Seizures inconvenience? Policy, discretion accidental discoveries policing Norwegian border.Crime Law Soc. Change. 64: 177-192Crossref (16) ScholarBox 1A Glossary TermsBias: systematic (as opposed random) causing loss accuracy precision).Big data: too traditional data-handling software, well highly variable data. require new methods storage handle volumes tease signal noise.Citizen intentional, voluntary amateur enthusiasts research activities. Participants provide (observational experimental) facilities researchers also input project design.Crowd sensing: numbers individuals, each submits (usually) web-enabled mobile devices smartphones.Distributed mind: describing task split numerous individuals same time, e.g., protein-folding project, foldit (https://fold.it).Gamification: game-design game principles non-game contexts.Observation process: factors lead event being recorded observation. people are, chance detect event, motivation record record.Occupancy modeling: analytical explicitly (probability detection) event), two otherwise confounded. Typically analyzes binary occurrence repeat samples, although extensions allow different structures.Semi-structured comprising observations made without standardized protocol, important metadata regarding process.Unstructured protocol.Web scraping: extraction amounts sources, occur knowledge permission content creator.Whole-system approach: method conducting question formulated investigated explicit consideration full context phenomena interest, analysis, responses interested parties occur. Bias: precision). noise. Citizen design. Crowd smartphones. Distributed (https://fold.it). Gamification: contexts. Observation record. Occupancy structures. Semi-structured process. Unstructured protocol. Web creator. Whole-system Messy advantages structured surveys, low cost, easy accessibility, high volume, real-world relevance. In cases, only source interest. past abundance distribution organism impossible reference museum historical Seebens al.24Seebens H. Blackburn Dyer E.E. Genovesi Jeschke Pagad Pyšek Winter Arianoutsou No saturation accumulation alien species worldwide.Nat. 8: 14435Crossref (716) McClenachan al.25McClenachan McKenzie M.G. Drew J.A. surprising results best practices ecology.BioScience. 65: 932-939Crossref (41) Scholar). working purposes web-scraping listings products offered sale online) allows activities putting themselves physical danger. limitations. Any poses three main challenge: accounting errors mistakes incorrect identification); random variation "noise"), inherent observed; (3) observer bias—systematic arising preferential recording certain events). especially pervasive, requiring particularly careful consideration.26Isaac N.J. Pocock Bias biological records.Biol. Linn. 115: 522-531Crossref (136) 27Isaac van Strien A.J. August de Zeeuw M.P. Anderson B. Statistics extracting signals noisy data.Methods 5: 1052-1060Crossref (226) 28Aceves-Bueno Adeleye Feraud Huang Y. Tao Yang S.E. quantitative review.Bull. Am. 98: 278-290Crossref upon insights natural synthesize offer guidance wishing engage particular, challenge notion after collection. discuss weighing, early stage, disadvantages using against user-designed, scientifically survey. lay steps appraise candidate beginning underlying process—specifically, way affected motivations, needs, backgrounds observers. illustrate exercise serves purposes: anticipating identifying align users mutual benefit. Finally, argue realizing requires practitioners adopt whole-system entire life cycle, problem formulation presentation results. reach internet, coupled rapid uptake devices, created unparalleled gather low-cost various types.29McKinley D.C. Miller-Rushing Brown Cook-Patton S.C. Evans D.M. French Parrish J.K. science, resource management, protection.Biol. 15-28Crossref (335) Scholar,30Sheil Mugerwa Fegraus E.H. African golden cats, serendipity: tapping camera trap revolution.South Afr. Wildl. Res. 2013; 43: 74-78Crossref (5) acquire, subsequent cost collation, appropriate interpretation terms both money; always worth using. To take example volunteer-collected organizations managing substantial funding attract, retain, support volunteers; maintain data-entry systems; validate data.31Pocock D.S. Sheppard L.J. H.E. Choosing Using Science: A Guide When How Use Science Monitor Biodiversity Environment. NERC/Centre Ecology & Hydrology, 2014Google al.31Pocock useful flowchart volunteer through costs benefits indicating whether feasible. UK projects, Breeding Bird Survey Butterfly Monitoring Scheme,19Brereton Botham Middlebrook I. Randle Z. Noble United Kingdom Scheme Annual Report 2016. Centre Hydrology Conservation, 2017https://www.ukbms.org/docs/reports/2016/Butterfly%20Ann%20Report%202016.pdfGoogle Scholar,32Harris Massimino Gillings Eaton Balmer Procter Pearce-Higgins Woodcock 2017. British Trust Ornithology, 2018Google directly inform policy providing indicators) £70,000 £150,000 annually maintain.33Roy Preston C.D. Savage Tweddle Robinson Understanding Environmental Monitoring: Final Behalf Framework. NERC Hydrology/Natural History Museum, 2012Google rely observers, sampling times, locations, protocols nonetheless carefully planned, making amenable therefore representing good value money. By less money support, information; DOFbasen, contains sightings birds Denmark, showed it detected fewer than half declines rates Denmark apparent more Danish Common Scheme.34Kamp Oppel Heldbjerg Nyegaard Donald P.F. fail common Denmark.Divers. Distrib. 22: 1024-1035Crossref (60) similar comparison consistent agreement calculated simple techniques (90 141 species' positively correlated sets), was widespread species.35Boersch-Supan P.H. Trask A.E. Baillie Robustness avian trend species-dependent.Biol. 240: 108286Crossref (15) There corollary set messiness produced engagement direct participation.18Ballard Additionally, yield might unanticipated challenges36Kelling Riedewald Caruana Hooker paradigm studies.BioScience. 2009; 59: 613-620Crossref (205) (although exclusive type data). Partridge Count Scheme, scheme up Game Wildlife Conservation monitor gray partridge (Perdix perdix) breeding success, subsequently used evaluate agri-environmental schemes.37Ewald Aebischer Richardson Grice Cooke agri-environment schemes grey partridges farm level England.Agric. Ecosyst. 2010; 138: 55-63Crossref (32) Many published studies now include raw supplementary material, sharing easier. It authors code able existing cleaning processing. another posts platforms Twitter searched warnings biosecurity risks agricultural pests;38Caley Welvaert surveillance: estimating reporting probabilities insects concern.J. Pest 93: 543-550Crossref (9) conduct surveillance professional would vastly higher necessarily better information. Social illuminate clandestine behaviors trading (IWT) when pertinent happening, general its characteristics rather detailed questions absolute magnitudes.22Hinsley most pervasive forms offering very little scope mitigation. case IWT, privacy settings carried forums, open Facebook groups, vendors will advertise closed, private groups sell messages. know what proportion recorded. Furthermore, openly, sales usually place private, meaning location identity consumer, even final price agreed, known. Careful framing questions, together limitations, employed before datasets.22Hinsley Datasets originate postings bias, result (which conducted researchers). Data collected non-researcher answer questions; examples harvest hunters develop strategies area management.39Critchlow Plumptre Alidria Nsubuga Driciru Rwetsiba Wanyama Beale C.M. Improving law-enforcement effectiveness efficiency areas ranger-collected data.Conserv. 572-580Crossref (42) Scholar,40Bischof Nilsen E.B. Brøseth Männil Ozoliņš Linnell Implementation uncertainty recreational manage carnivores.J. 824-832Crossref nothing acquire expensive use. situations, collected, so countered during phase. Indeed, statistics driven part requirement large, datasets.41Gandomi Haider Beyond hype: concepts, methods, analytics.Int. Inf. 137-144Crossref (1807) Complex enabled countless sciences greatly utility implications. Firstly, specialist analysts software computing Bayesian analysis), barriers ongoing expertise); open-source R should, however, increase accessibility analysis. Secondly, statistically characterizing clear covariate unavailable. free choose times places cannot accounted standardization could achieved strict, formalized protocol), availability important. gathered post hoc users. Thirdly, greater sophistication summarize non-specialist audiences, one wish enthuse encourage communication outputs.42McInerny G.J. Chen Freeman Gavagha

Язык: Английский

---

Cyber–Physiochemical Interfaces

Advanced Materials, Год журнала: 2020, Номер 32(8)

Опубликована: Янв. 15, 2020

Abstract Living things rely on various physical, chemical, and biological interfaces, e.g., somatosensation, olfactory/gustatory perception, nervous system response. They help organisms to perceive the world, adapt their surroundings, maintain internal external balance. Interfacial information exchanges are complicated but efficient, delicate precise, multimodal unisonous, which has driven researchers study science of such interfaces develop techniques with potential applications in health monitoring, smart robotics, future wearable devices, cyber physical/human systems. To understand better issues these a cyber–physiochemical interface (CPI) that is capable extracting biophysical biochemical signals, closely relating them electronic, communication, computing technology, provide core for aforementioned applications, proposed. The scientific technical progress CPI summarized, challenges strategies building stable including materials, sensor development, integration, data processing discussed. It hoped this will result an unprecedented multi‐disciplinary network collaboration explore much uncharted territory progress, providing inspiration—to development next‐generation personal healthcare sports‐technology, adaptive prosthetics augmentation human capability, etc.

Язык: Английский

---