FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4221127087> ?p ?o ?g. }

• W4221127087 endingPage "e548" @default.
• W4221127087 startingPage "e547" @default.
• W4221127087 abstract "There are several emerging AI technologies that aim to enhance surgical care pathways over the coming decade. In particular, these are related to (1) diagnostics, (2) pre-operative planning, (3) intra-operative guidance and (4) surgical robotics.1 This trend has been mirrored bn the sharp increase in the number of surgical studies evaluating the use of AI. Despite this fervour, very few AI devices have reached the point of clinical implementation within surgical environments.2 This disconnect between ‘in silico bench’ and ‘bedside’ is a multifaceted issue related to technological, regulatory, and economic factors. However, this divide is also exacerbated by the variable quality of study reporting in this field; an issue perpetuated by the absence of AI-specific reporting guidelines for both pre-clinical and clinical AI studies. Poor reporting has been shown to hinder the clinical translation of otherwise promising research findings.3 One of the principle means of mitigating the risk of poor reporting is through adherence to consensus reporting standards. These tools, many of which are endorsed across biomedical journals, describe the critical information expected in research manuscripts. They typically consist of a checklist of minimally essential items, a flow diagram, and are accompanied with a longer elaboration and explanation document consisting of rationale and examples of good reporting. The enhancing the quality and transparency of health research network4 hosts a comprehensive library of reporting guidelines, with prominent examples including standard protocol items: recommendations for interventional trials (SPIRIT), consolidated standards of reporting trials (CONSORT), standards for reporting diagnostic accuracy (STARD), and Transparent Reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD). A caveat to the use of such instruments is that they are necessarily study design specific. When a study design specific tool is not available, extensions to existing tools are actively encouraged. In the absence of AI specific tools, researchers and readers have been left with the after reporting pitfalls (Supplemental Digital Content Table 1, https://links.lww.com/SLA/D514). If left unchecked, poorly reported studies can have a direct and deleterious impact upon how surgical departments shape their service. As a practical example, when reporting upon AI devices that can diagnose cancer from screening mammograms, studies should look to specify upon how the device demonstrates equitable performance across population groups, whether the system fits within existing clinical pathways, against what reference standard performance was measured against and, if it is an adaptive device, how performance alters with dataset shifts. Failing to mention this can lead to the adoption of expensive and impractical technologies that can potentially exacerbate health inequalities through the skewed diagnosis of target pathologies. Therefore, to fill this reporting guideline void, AI extensions are being undertaken for the following: 1) SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) 2013 2) CONSORT (Consolidated Standards of Reporting Trials) 2010 3) STARD (Standards for Reporting of Diagnostic Accuracy Studies) 2015 4) TRIPOD (Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis) 2015 The first in these series of reporting guidelines aim to improve the reporting of clinical trial protocols (SPIRIT-AI5) and as well as the eventual findings of the trials themselves (CONSORT-AI6). In combination, these documents are designed to enable readers to (1) understand the background, rationale, population, methods, statistical analyses, and ethico-legal considerations, (2) assist replication of key aspects of the trial (including implementation of the intervention) and (3) assist in the appraisal the study's scientific rigour. To achieve these goals for medical AI research, SPIRIT-AI and CONSORT-AI contain 15 and 14 new AI specific items respectively. These items are centred around 4 general themes, which tackle the issues noted in Supplemental Digital Content Table 1, https://links.lww.com/SLA/D514: (1) AI model, (2) dataset, (3) infrastructure and (4) transparency. 1) AI model: Firstly, there need to be details regarding the development and any subsequent validation of the AI model, as this is often underreported in clinical validation studies. Thereafter, a clear use case needs to be defined in addition to information related to how the model will fit into existing clinical workflows and interact with human end-users.7 SPIRIT-AI and CONSORT-AI highlight the importance for clear criteria as to who (experience and skill level) interprets the output as well as how to interpret the output (e.g., diagnostic probability or classification) in the context of making a clinical decision. 2) Dataset: Eligibility criteria at both a participant and data level; wisely kept as separate items, should be reported. For example, there should be clearly stated minimum input data entry requirements, which can be related to image resolution or data format. Moreover, clear procedures regarding acquisition, selection, and pre-processing should be described to allow input standardisation across study sites. 3) Infrastructure: Given the complexity of the AI systems proposed, several reporting items are dedicated to highlighting the specific software, hardware and training requirements that are required for the intervention to be incorporated. These requirements may be divided into both on-site and off-site requirements; both of which are vital to delineate when appraisal the deployment of such interventions. 4) Transparency: Both SPIRIT-AI and CONSORT-AI highlight that transparency is an essential consideration across the study reporting process. There are items which encourage specifying both the presence and intended use of an AI system in either the title or the abstract, which allows readers to understand the intended use of the AI intervention in an unambiguous manner. Authors are encouraged to state the version number of the AI model associated with the study, with any changes to this requiring clear justification. There is a recommendation to “describe results of any analysis of performance errors and how errors were identified, where applicable”. This is particularly warranted as regulatory bodies note that adverse incidents emerging from diagnostic AI as a medical device are substantially under reported and are difficult to capture as much of the harm to users is indirect. Lastly, there are items related to how the AI model and/or its code can be accessed, including any restrictions to access or re-use. In addition to SPIRIT-AI and CONSORT-AI, further specific guidance related to the reporting of diagnostic accuracy studies and multivariable prediction studies will be provided by the forthcoming STARD-AI8 and TRIPOD-AI9 initiatives, which will be released in late 2021. These harmonised initiatives will ensure that there is AI specific reporting guideline coverage of the entire research pipeline, from the construction of a study protocol through to the reporting of trial results of various methodologies. These instruments will have a significant and direct impact upon the quality and reproducibility of surgical research (e.g. studies assessing computer vision which report upon intraoperative guidance10 and surgical safety11) in addition to critical research in partner specialties (e.g. applications within computational pathology12 and radiomics13 that aim to drastically improve the diagnosis, treatment and prognostication of cancer that is amenable to surgical resection). On a wider systems level, these instruments will also serve to maximise (1) the clinical and cost utility of research being conducted, (2) assist regulators, policy makers, funders, clinicians, and scientists in approval processes as well as ultimately (3) drive improved outcomes for patients. Given that surgeons assume responsibility for many of these roles, it is vital that all surgical stakeholders, ranging from those who intend to undertake, appraise, fund, or even utilise AI research findings, are aware of these vital developments in the field. For those who wish to collaborate with us, as members of the study groups undertaking the STARD-AI and TRIPOD-AI initiatives, we would welcome the opportunity to work together to help drive the integration of these technologies into clinical care." @default.
• W4221127087 created "2022-04-03" @default.
• W4221127087 creator A5003509342 @default.
• W4221127087 creator A5018384194 @default.
• W4221127087 creator A5029868483 @default.
• W4221127087 creator A5043205117 @default.
• W4221127087 creator A5053747302 @default.
• W4221127087 creator A5055785355 @default.
• W4221127087 creator A5058163838 @default.
• W4221127087 creator A5061726857 @default.
• W4221127087 date "2021-11-11" @default.
• W4221127087 modified "2023-09-24" @default.
• W4221127087 title "Developing Specific Reporting Standards in Artificial Intelligence Centred Research" @default.
• W4221127087 cites W2937186836 @default.
• W4221127087 cites W3033545112 @default.
• W4221127087 cites W3045477517 @default.
• W4221127087 cites W3112363719 @default.
• W4221127087 cites W3119083236 @default.
• W4221127087 cites W3134491265 @default.
• W4221127087 cites W3136176646 @default.
• W4221127087 cites W3136552952 @default.
• W4221127087 cites W3136960503 @default.
• W4221127087 doi "https://doi.org/10.1097/sla.0000000000005294" @default.
• W4221127087 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/35120063" @default.
• W4221127087 hasPublicationYear "2021" @default.
• W4221127087 type Work @default.
• W4221127087 citedByCount "5" @default.
• W4221127087 countsByYear W42211270872023 @default.
• W4221127087 crossrefType "journal-article" @default.
• W4221127087 hasAuthorship W4221127087A5003509342 @default.
• W4221127087 hasAuthorship W4221127087A5018384194 @default.
• W4221127087 hasAuthorship W4221127087A5029868483 @default.
• W4221127087 hasAuthorship W4221127087A5043205117 @default.
• W4221127087 hasAuthorship W4221127087A5053747302 @default.
• W4221127087 hasAuthorship W4221127087A5055785355 @default.
• W4221127087 hasAuthorship W4221127087A5058163838 @default.
• W4221127087 hasAuthorship W4221127087A5061726857 @default.
• W4221127087 hasBestOaLocation W42211270871 @default.
• W4221127087 hasConcept C17744445 @default.
• W4221127087 hasConcept C19527891 @default.
• W4221127087 hasConcept C199539241 @default.
• W4221127087 hasConcept C2522767166 @default.
• W4221127087 hasConcept C2779473830 @default.
• W4221127087 hasConcept C41008148 @default.
• W4221127087 hasConcept C71924100 @default.
• W4221127087 hasConceptScore W4221127087C17744445 @default.
• W4221127087 hasConceptScore W4221127087C19527891 @default.
• W4221127087 hasConceptScore W4221127087C199539241 @default.
• W4221127087 hasConceptScore W4221127087C2522767166 @default.
• W4221127087 hasConceptScore W4221127087C2779473830 @default.
• W4221127087 hasConceptScore W4221127087C41008148 @default.
• W4221127087 hasConceptScore W4221127087C71924100 @default.
• W4221127087 hasIssue "3" @default.
• W4221127087 hasLocation W42211270871 @default.
• W4221127087 hasLocation W42211270872 @default.
• W4221127087 hasLocation W42211270873 @default.
• W4221127087 hasOpenAccess W4221127087 @default.
• W4221127087 hasPrimaryLocation W42211270871 @default.
• W4221127087 hasRelatedWork W2104151291 @default.
• W4221127087 hasRelatedWork W2140378881 @default.
• W4221127087 hasRelatedWork W2319790315 @default.
• W4221127087 hasRelatedWork W2748952813 @default.
• W4221127087 hasRelatedWork W2899084033 @default.
• W4221127087 hasRelatedWork W3031052312 @default.
• W4221127087 hasRelatedWork W3113343617 @default.
• W4221127087 hasRelatedWork W3167030926 @default.
• W4221127087 hasRelatedWork W4220654564 @default.
• W4221127087 hasRelatedWork W4381249388 @default.
• W4221127087 hasVolume "275" @default.
• W4221127087 isParatext "false" @default.
• W4221127087 isRetracted "false" @default.
• W4221127087 workType "article" @default.