The Impact of Artificial Intelligence on Health Equity in Oncology: Scoping Review DOI Creative Commons
Paul Istasy, Wen Shen Lee,

Alla Iansavichene

et al.

Journal of Medical Internet Research, Journal Year: 2022, Volume and Issue: 24(11), P. e39748 - e39748

Published: Aug. 25, 2022

Background The field of oncology is at the forefront advances in artificial intelligence (AI) health care, providing an opportunity to examine early integration these technologies clinical research and patient care. Hope that AI will revolutionize care delivery improve outcomes has been accompanied by concerns about impact on equity. Objective We aimed conduct a scoping review literature address question, “What are current potential impacts equity oncology?” Methods Following PRISMA-ScR (Preferred Reporting Items for Systematic Reviews Meta-Analyses extension Scoping Reviews) guidelines reviews, we systematically searched MEDLINE Embase electronic databases from January 2000 August 2021 records engaging with key concepts AI, equity, oncology. included all English-language articles engaged 3 concepts. Articles were analyzed qualitatively themes pertaining influence Results Of 14,011 records, 133 (0.95%) identified our included. general literature: use reduce disparities (58/133, 43.6%), surrounding bias (16/133, 12.1%), biological social determinants (55/133, 41.4%). A total 3% (4/133) focused many themes. Conclusions Our revealed main oncology, which relate AI’s ability help disparities, its mitigate or exacerbate bias, capability elucidate health. Gaps lack discussion ethical challenges application low- middle-income countries, problems algorithms, justification over traditional statistical methods specific questions highlights need gaps ensure more equitable cancer practice. limitations study include exploratory nature, focus as opposed sectors, analysis solely articles.

Language: Английский

Artificial Intelligence (AI): Multidisciplinary perspectives on emerging challenges, opportunities, and agenda for research, practice and policy DOI
Yogesh K. Dwivedi, Laurie Hughes, Elvira Ismagilova

et al.

International Journal of Information Management, Journal Year: 2019, Volume and Issue: 57, P. 101994 - 101994

Published: Aug. 27, 2019

Language: Английский

Citations

2550

Key challenges for delivering clinical impact with artificial intelligence DOI Creative Commons
Christopher Kelly, Alan Karthikesalingam, Mustafa Suleyman

et al.

BMC Medicine, Journal Year: 2019, Volume and Issue: 17(1)

Published: Oct. 29, 2019

Abstract Background Artificial intelligence (AI) research in healthcare is accelerating rapidly, with potential applications being demonstrated across various domains of medicine. However, there are currently limited examples such techniques successfully deployed into clinical practice. This article explores the main challenges and limitations AI healthcare, considers steps required to translate these potentially transformative technologies from Main body Key for translation systems include those intrinsic science machine learning, logistical difficulties implementation, consideration barriers adoption as well necessary sociocultural or pathway changes. Robust peer-reviewed evaluation part randomised controlled trials should be viewed gold standard evidence generation, but conducting practice may not always appropriate feasible. Performance metrics aim capture real applicability understandable intended users. Regulation that balances pace innovation harm, alongside thoughtful post-market surveillance, ensure patients exposed dangerous interventions nor deprived access beneficial innovations. Mechanisms enable direct comparisons must developed, including use independent, local representative test sets. Developers algorithms vigilant dangers, dataset shift, accidental fitting confounders, unintended discriminatory bias, generalisation new populations, negative consequences on health outcomes. Conclusion The safe timely clinically validated appropriately regulated can benefit everyone challenging. evaluation, using intuitive clinicians ideally go beyond measures technical accuracy quality care patient outcomes, essential. Further work (1) identify themes algorithmic bias unfairness while developing mitigations address these, (2) reduce brittleness improve generalisability, (3) develop methods improved interpretability learning predictions. If goals achieved, benefits likely transformational.

Language: Английский

Citations

1621

On the Opportunities and Risks of Foundation Models DOI Creative Commons
Rishi Bommasani,

Drew A. Hudson,

Ehsan Adeli

et al.

arXiv (Cornell University), Journal Year: 2021, Volume and Issue: unknown

Published: Jan. 1, 2021

AI is undergoing a paradigm shift with the rise of models (e.g., BERT, DALL-E, GPT-3) that are trained on broad data at scale and adaptable to wide range downstream tasks. We call these foundation underscore their critically central yet incomplete character. This report provides thorough account opportunities risks models, ranging from capabilities language, vision, robotics, reasoning, human interaction) technical principles(e.g., model architectures, training procedures, data, systems, security, evaluation, theory) applications law, healthcare, education) societal impact inequity, misuse, economic environmental impact, legal ethical considerations). Though based standard deep learning transfer learning, results in new emergent capabilities,and effectiveness across so many tasks incentivizes homogenization. Homogenization powerful leverage but demands caution, as defects inherited by all adapted downstream. Despite impending widespread deployment we currently lack clear understanding how they work, when fail, what even capable due properties. To tackle questions, believe much critical research will require interdisciplinary collaboration commensurate fundamentally sociotechnical nature.

Language: Английский

Citations

1591

The multi-factorial nature of clinical multidrug resistance in cancer DOI
Yehuda G. Assaraf, Anamaria Brozović, Ana Cristina Gonçalves

et al.

Drug Resistance Updates, Journal Year: 2019, Volume and Issue: 46, P. 100645 - 100645

Published: Sept. 1, 2019

Language: Английский

Citations

446

Use of AI-based tools for healthcare purposes: a survey study from consumers’ perspectives DOI Creative Commons
Pouyan Esmaeilzadeh

BMC Medical Informatics and Decision Making, Journal Year: 2020, Volume and Issue: 20(1)

Published: July 22, 2020

Abstract Background Several studies highlight the effects of artificial intelligence (AI) systems on healthcare delivery. AI-based tools may improve prognosis, diagnostics, and care planning. It is believed that AI will be an integral part services in near future incorporated into several aspects clinical care. Thus, many technology companies governmental projects have invested producing medical applications. Patients can one most important beneficiaries users applications whose perceptions affect widespread use tools. should ensured they not harmed by devices, instead, benefited using for purposes. Although enhance outcomes, possible dimensions concerns risks addressed before its integration with routine Methods We develop a model mainly based value due to specificity field. This study aims at examining perceived benefits devices decision support (CDS) features from consumers’ perspectives. online survey collect data 307 individuals United States. Results The proposed identifies sources motivation pressure patients development devices. results show technological, ethical (trust factors), regulatory significantly contribute healthcare. Of three categories, technological (i.e., performance communication feature) are found significant predictors risk beliefs. Conclusions sheds more light factors affecting proposes some recommendations how practically reduce these concerns. findings this provide implications research practice area CDS. Regulatory agencies, cooperation institutions, establish normative standard evaluation guidelines implementation Regular audits ongoing monitoring reporting used continuously evaluate safety, quality, transparency, services.

Language: Английский

Citations

342

Artificial intelligence (AI) and big data in cancer and precision oncology DOI Creative Commons
Zodwa Dlamini, Flavia Zita Francies, Rodney Hull

et al.

Computational and Structural Biotechnology Journal, Journal Year: 2020, Volume and Issue: 18, P. 2300 - 2311

Published: Jan. 1, 2020

Artificial intelligence (AI) and machine learning have significantly influenced many facets of the healthcare sector. Advancement in technology has paved way for analysis big datasets a cost- time-effective manner. Clinical oncology research are reaping benefits AI. The burden cancer is global phenomenon. Efforts to reduce mortality rates requires early diagnosis effective therapeutic interventions. However, metastatic recurrent cancers evolve acquire drug resistance. It imperative detect novel biomarkers that induce resistance identify targets enhance treatment regimes. introduction next generation sequencing (NGS) platforms address these demands, revolutionised future precision oncology. NGS offers several clinical applications important risk predictor, detection disease, by medical imaging, accurate prognosis, biomarker identification discovery. generates large demand specialised bioinformatics resources analyse data relevant clinically significant. Through AI, diagnostics prognostic prediction enhanced with imaging delivers high resolution images. Regardless improvements technology, AI some challenges limitations, application remains be validated. By continuing progression innovation show great promise.

Language: Английский

Citations

274

The ethical, legal and social implications of using artificial intelligence systems in breast cancer care DOI Open Access
Stacy M. Carter, Wendy Rogers, Khin Than Win

et al.

The Breast, Journal Year: 2019, Volume and Issue: 49, P. 25 - 32

Published: Oct. 11, 2019

Breast cancer care is a leading area for development of artificial intelligence (AI), with applications including screening and diagnosis, risk calculation, prognostication clinical decision-support, management planning, precision medicine. We review the ethical, legal social implications these developments. consider values encoded in algorithms, need to evaluate outcomes, issues bias transferability, data ownership, confidentiality consent, legal, moral professional responsibility. potential effects patients, on trust healthcare, provide some science explanations apparent rush implement AI solutions. conclude by anticipating future directions breast care. Stakeholders healthcare should acknowledge that their enterprise an challenge, not just technical challenge. Taking challenges seriously will require broad engagement, imposition conditions implementation, pre-emptive systems oversight ensure does run ahead evaluation deliberation. Once becomes institutionalised, it may be difficult reverse: proactive role government, regulators groups help introduction robust research contexts, sound evidence base regarding real-world effectiveness. Detailed public discussion required what kind acceptable rather than simply accepting offered, thus optimising outcomes health systems, professionals, society those receiving

Language: Английский

Citations

236

Current status and applications of Artificial Intelligence (AI) in medical field: An overview DOI
Abid Haleem, Mohd Javaid, Ibrahim Haleem Khan

et al.

Current Medicine Research and Practice, Journal Year: 2019, Volume and Issue: 9(6), P. 231 - 237

Published: Nov. 1, 2019

Language: Английский

Citations

187

M3DISEEN: A novel machine learning approach for predicting the 3D printability of medicines DOI
Moe Elbadawi,

Brais Muñiz Castro,

Francesca K. H. Gavins

et al.

International Journal of Pharmaceutics, Journal Year: 2020, Volume and Issue: 590, P. 119837 - 119837

Published: Sept. 20, 2020

Language: Английский

Citations

180

Patients’ Perceptions Toward Human–Artificial Intelligence Interaction in Health Care: Experimental Study DOI Creative Commons
Pouyan Esmaeilzadeh, Tala Mirzaei, Spurthy Dharanikota

et al.

Journal of Medical Internet Research, Journal Year: 2021, Volume and Issue: 23(11), P. e25856 - e25856

Published: Nov. 25, 2021

It is believed that artificial intelligence (AI) will be an integral part of health care services in the near future and incorporated into several aspects clinical such as prognosis, diagnostics, planning. Thus, many technology companies have invested producing AI applications. Patients are one most important beneficiaries who potentially interact with these technologies applications; thus, patients' perceptions may affect widespread use AI. should ensured applications not harm them, they instead benefit from using for purposes. Although human-AI interaction can enhance outcomes, possible dimensions concerns risks addressed before its integration routine care.The main objective this study was to examine how potential users (patients) perceive benefits, risks, their purposes different if faced three service encounter scenarios.We designed a 2×3 experiment crossed type condition (ie, acute or chronic) types encounters between patients physicians substituting technology, augmenting no traditional in-person visit). We used online survey collect data 634 individuals United States.The interactions conditions significantly influenced individuals' privacy concerns, trust issues, communication barriers, about transparency regulatory standards, liability intention across six scenarios. found significant differences among scenarios regarding performance risk social biases.The results imply incompatibility instrumental, technical, ethical, values reason rejecting care. there still various associated implementing diagnostics treatment recommendations both chronic illnesses. The also evident recommendation system under physician experience, wisdom, control. Prior rollout AI, more studies needed identify challenges raise This could provide researchers managers critical insights determinants Regulatory agencies establish normative standards evaluation guidelines cooperation institutions. Regular audits ongoing monitoring reporting systems continuously evaluate safety, quality, transparency, ethical factors

Language: Английский

Citations

128