FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3198086504> ?p ?o ?g. }

• W3198086504 abstract "Genetic information is an ethically salient mediator of medical intervention due to its potential to improve health. However, individuals in low-income and middle-income countries (LMICs) might have limited access to genetic services. Artificial intelligence has the potential to improve health equity and facilitate access to genetic expertise in areas where medical services are unavailable, unreliable, or unaffordable. With the aim of improving access in places where such barriers are faced, Antonio Porras and colleagues,1Porras AR Rosenbaum K Tor-Diez C Summar M Linguraru MG Development and evaluation of a machine learning-based point-of-care screening tool for genetic syndromes in children: a multinational retrospective study.Lancet Digit Health. 2021; (published online Sept 1.)https://doi.org/10.1016/S2589-7500(21)00137-0Summary Full Text Full Text PDF PubMed Scopus (3) Google Scholar present a novel facial phenotyping technology to screen children and adolescents for genetic syndromes in The Lancet Digital Health. Intended for use at the point-of-care, the tool predicts the likelihood of a patient having an underlying genetic disorder based on a machine learning-enabled analysis of their facial features. It does not offer suggested diagnoses, unlike similar tools, but is intended for use as a screening tool to efficiently allocate patients for genetics consultation. Based on a dataset of 2800 facial photographs of children and adolescents (1400 images of children with 128 different genetic conditions and 1400 images of controls matched by age, sex, and race or ethnicity), the deep learning-based model had an accuracy of 88% (95% CI 87–89) for the detection of a genetic syndrome, with 90% sensitivity (95% CI 88–92) and 86% specificity (95% CI 84–88). Although the performance of the tool did not differ by sex or age, accuracy was greater in White (90%, 89–91) and Hispanic populations (91%, 88–94) than in African (84%, 81–87) and Asian populations (82%, 78–86). Porras and colleagues note1Porras AR Rosenbaum K Tor-Diez C Summar M Linguraru MG Development and evaluation of a machine learning-based point-of-care screening tool for genetic syndromes in children: a multinational retrospective study.Lancet Digit Health. 2021; (published online Sept 1.)https://doi.org/10.1016/S2589-7500(21)00137-0Summary Full Text Full Text PDF PubMed Scopus (3) Google Scholar that an important next step involves a prospective clinical trial. As for other diagnostic-assistive technologies,2Keane PA Topol EJ With an eye to AI and autonomous diagnosis.NPJ Digit Med. 2018; 1: 40Crossref PubMed Google Scholar an appropriately designed prospective trial measuring patient-centred outcomes will provide reliable evidence of whether the tool is not just accurate, but also beneficial. Two aims are relevant; namely, (1) prospective demonstration of the validity of the tool by accurately capturing cases later confirmed to have genetic diagnoses; and (2) comparison against the status quo by demonstrating the relative advantage of the tool over clinicians' own gestalts in identifying a child with facial dysmorphology. Evaluating against the status quo is central to determining clinical value. LMICs are not a monolith. As such, a benefit identified in one region might not apply equally to another. Transparent reporting of clinical trials by specifying the conditions under which the use of the model was evaluated and shown to be of value can guide its efficient translation into practice.3Liu X Rivera SC Faes L et al.Reporting guidelines for clinical trials evaluating artificial intelligence interventions are needed.Nat Med. 2019; 25: 1467-1468Crossref PubMed Scopus (43) Google Scholar, 4Rivera S Liu X Chan AW et al.Guidelines for clinical trial protocols for interventions involving artificial intelligence: the SPIRIT-AI extension.Nat Med. 2020; 26: 1351-1363Crossref PubMed Scopus (63) Google Scholar These reported conditions include features of the care pathway that are crucial for the technology to have a positive impact on patients' access to and benefit from genetics services. For example, in some LMICs, the implementation of such tools1Porras AR Rosenbaum K Tor-Diez C Summar M Linguraru MG Development and evaluation of a machine learning-based point-of-care screening tool for genetic syndromes in children: a multinational retrospective study.Lancet Digit Health. 2021; (published online Sept 1.)https://doi.org/10.1016/S2589-7500(21)00137-0Summary Full Text Full Text PDF PubMed Scopus (3) Google Scholar could provide motivation to create novel care pathways (eg, telehealth services), enabling connectivity between under-served areas and genetics services. However, the cost of expert assessment is a critical barrier to patients being able to effectively access genetics in LMICs that can blunt the potential benefit of such screening tools and hamper efforts toward equity. Identification of patients at risk of having a genetic condition might not have the intended positive impact without the means to afford diagnostic tests, treatment, or supportive care, which could introduce an under-appreciated potential harm to families. Porras and colleagues recognise the marked disparities in access and diagnostic accuracy across racial and ethnic groups—accordingly, they tested the performance of their model along a metric of racial or ethnic background.1Porras AR Rosenbaum K Tor-Diez C Summar M Linguraru MG Development and evaluation of a machine learning-based point-of-care screening tool for genetic syndromes in children: a multinational retrospective study.Lancet Digit Health. 2021; (published online Sept 1.)https://doi.org/10.1016/S2589-7500(21)00137-0Summary Full Text Full Text PDF PubMed Scopus (3) Google Scholar A clear understanding of causal reasons for these disparities is important.5McCradden MD Joshi S Mazwi M Anderson JA Ethical limitations of algorithmic fairness solutions in health care machine learning.Lancet Digit Health. 2020; 2: e221-e223Summary Full Text Full Text PDF PubMed Scopus (21) Google Scholar Ethnicity is an important factor in genetics, but it does not necessarily connote the patterns of social disadvantage that are of relevance to promoting health equity. By contrast, race quite often betrays patterns of privilege and disadvantage, which might be linked with the likelihood of receiving a genetic diagnosis, access to genetics services, and clinical outcomes.6Hall M Olopade OI Confronting genetic testing disparities: knowledge is power.JAMA. 2005; 293: 1783-1785Crossref PubMed Scopus (69) Google Scholar A well performing algorithm is the first step; but to promote equity, this performance must generate accountability, ensuring that all children across the spectrum of genetic conditions (not just those with a condition a model is trained on) can benefit from genetic knowledge to improve their health outcomes. Importantly, the authors do not suggest that the tool could be used as a wholescale replacement for newborn screening, nor for use in isolation from local clinical teams, health-care systems, and social services. We echo these comments and advise caution. Most genetic and metabolic conditions included in universal newborn screening programmes do not have discernible dysmorphic features and would not be identified with this type of screening aid.7Watson MS Mann MY Lloyd-Puryear MA Rinaldo R Howell RR Newborn screening: toward a uniform screening panel and system.Genet Med. 2006; 8: 1-252SSummary Full Text Full Text PDF PubMed Scopus (13) Google Scholar A negative result on such a screening tool might also provide a false sense of security for genetic conditions that do not have identifiable facial morphological features. Further, screening for conditions that might have no treatment, or for which natural history is poorly understood might not be desirable.8Dobrow MJ Hagens V Chafe R Sullivan T Rabeneck L Consolidated principles for screening based on a systematic review and consensus process.CMAJ. 2018; 190: E422-E429Crossref PubMed Scopus (65) Google Scholar Unresolved issues related to children's privacy, data security, and surrogate consent for facial image analysis with machine learning9McCradden MD Patel E Chad L The point-of-care use of a facial phenotyping tool in the genetics clinic: an ethics tête-a-tête.Am J Med Genet. 2021; 185: 658-660Crossref PubMed Scopus (2) Google Scholar could also be magnified when applied at scale. As outlined by Wilson and Jungner more than 50 years ago, more than accuracy is needed to justify implementation, including acceptability to the population, acceptable treatment is identified, and facilities for diagnosis and treatment should be available.8Dobrow MJ Hagens V Chafe R Sullivan T Rabeneck L Consolidated principles for screening based on a systematic review and consensus process.CMAJ. 2018; 190: E422-E429Crossref PubMed Scopus (65) Google Scholar, 10Wilson JMG Jungner G Principles and practice of screening for disease. World Health Organization, Geneva1968Google Scholar Ideally, communities and affected individuals (eg, disability rights groups) implicated in these technologies should participate in the design process to promote relevance and acceptability in their local context. We are encouraged by the efforts toward artificial intelligence that is equity focussed and commend Porras and colleagues'1Porras AR Rosenbaum K Tor-Diez C Summar M Linguraru MG Development and evaluation of a machine learning-based point-of-care screening tool for genetic syndromes in children: a multinational retrospective study.Lancet Digit Health. 2021; (published online Sept 1.)https://doi.org/10.1016/S2589-7500(21)00137-0Summary Full Text Full Text PDF PubMed Scopus (3) Google Scholar efforts to pursue a clinical trial to confirm the intended benefit of this technology. We declare no competing interests. Development and evaluation of a machine learning-based point-of-care screening tool for genetic syndromes in children: a multinational retrospective studyThis genetic screening technology could support early risk stratification at the point of care in global populations, which has the potential accelerate diagnosis and reduce mortality and morbidity through preventive care. Full-Text PDF Open Access" @default.
• W3198086504 created "2021-09-13" @default.
• W3198086504 creator A5000944814 @default.
• W3198086504 creator A5063981686 @default.
• W3198086504 date "2021-09-01" @default.
• W3198086504 modified "2023-09-23" @default.
• W3198086504 title "Screening for facial differences worldwide: equity and ethics" @default.
• W3198086504 cites W2062353292 @default.
• W3198086504 cites W2798221581 @default.
• W3198086504 cites W2888171911 @default.
• W3198086504 cites W3023069697 @default.
• W3198086504 cites W3109594159 @default.
• W3198086504 cites W3123582176 @default.
• W3198086504 cites W3136960503 @default.
• W3198086504 cites W3196566835 @default.
• W3198086504 cites W3202966610 @default.
• W3198086504 doi "https://doi.org/10.1016/s2589-7500(21)00179-5" @default.
• W3198086504 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/34481766" @default.
• W3198086504 hasPublicationYear "2021" @default.
• W3198086504 type Work @default.
• W3198086504 sameAs 3198086504 @default.
• W3198086504 citedByCount "0" @default.
• W3198086504 crossrefType "journal-article" @default.
• W3198086504 hasAuthorship W3198086504A5000944814 @default.
• W3198086504 hasAuthorship W3198086504A5063981686 @default.
• W3198086504 hasBestOaLocation W31980865041 @default.
• W3198086504 hasConcept C15744967 @default.
• W3198086504 hasConcept C17744445 @default.
• W3198086504 hasConcept C199539241 @default.
• W3198086504 hasConcept C199728807 @default.
• W3198086504 hasConceptScore W3198086504C15744967 @default.
• W3198086504 hasConceptScore W3198086504C17744445 @default.
• W3198086504 hasConceptScore W3198086504C199539241 @default.
• W3198086504 hasConceptScore W3198086504C199728807 @default.
• W3198086504 hasFunder F4320323654 @default.
• W3198086504 hasLocation W31980865041 @default.
• W3198086504 hasOpenAccess W3198086504 @default.
• W3198086504 hasPrimaryLocation W31980865041 @default.
• W3198086504 hasRelatedWork W1710126152 @default.
• W3198086504 hasRelatedWork W2053487507 @default.
• W3198086504 hasRelatedWork W2077865380 @default.
• W3198086504 hasRelatedWork W2134894512 @default.
• W3198086504 hasRelatedWork W2748952813 @default.
• W3198086504 hasRelatedWork W2755849213 @default.
• W3198086504 hasRelatedWork W2765597752 @default.
• W3198086504 hasRelatedWork W2899084033 @default.
• W3198086504 hasRelatedWork W2931662336 @default.
• W3198086504 hasRelatedWork W3049356719 @default.
• W3198086504 isParatext "false" @default.
• W3198086504 isRetracted "false" @default.
• W3198086504 magId "3198086504" @default.
• W3198086504 workType "article" @default.