echemi logo
Product
  • Product
  • Supplier
  • Inquiry
    Home > Medical News > Medical World News > Consensus standards for reporting AI interventions in clinical trials.

    Consensus standards for reporting AI interventions in clinical trials.

    • Last Update: 2020-10-25
    • Source: Internet
    • Author: User
    Search more information of high quality chemicals, good prices and reliable suppliers, visit www.echemi.com
    A new consensus statement, CONSORT-AI Extension, sets the rules for reporting clinical trials involving artificial intelligence (AI) interventions.
    , known as consORT-AI, is a new reporting guideline for evaluating clinical trials with AI-component interventions.
    Consensus Statement, published in Nature Medicine, was written by an international working group and includes 14 new projects for researchers to routinely incorporate AI interventions into their manuscripts when reporting them.
    statement called on researchers to report experiments that include AI, fully explain algorithm versions, input and output data, integrate into trial settings, consumer expertise, and protocols for action based on the recommendations of AI systems.
    idea is to promote transparent reporting of AI interventions, based on the checklist outlined in conSORT's 2010 statement, which provides minimal guidance for reporting randomized trials.
    conSORT 2010 statement was first introduced in 1996 and has been widely recognized in medical journals and updated over the past few years.
    CONSORT-AI extension was developed in cooperation with the accompanying statement of clinical trial scenarios, SPIRIT-AI (Standard Solutions Project: Intervention Test Recommendations - Artificial Intelligence).
    working group noted that "it has been recognized that the latest AI studies are not adequately reported and that existing reporting guidelines do not fully cover potential biases specific to AI systems."
    randomized controlled trials are welcome to seek evaluation of new interventions based on or including AI components, but at the same time there are concerns about the design and reporting of such trials.
    " list includes most of the clinical trial reports, from titles and summaries to funding disclosures.
    recommends that researchers indicate in the title or summary that the research intervention involves AI or machine learning, and specify the type of model.
    they should also describe the intended use of AI interventions in a clinical pathway, including their purpose and intended users (medical staff or patients).
    also report AI-specific information about the study participants, including participant-level inclusion and exclusion criteria and input data.
    , they should also describe how AI interventions can be integrated into a pilot environment, including any on-site or offs site requirements. The
    Consensus Statement also includes six requirements related to intervention information, around the version of the algorithm, how to obtain and select input data, how to evaluate and process poor quality and unavailable input data, whether human-computer interaction exists when processing input data, input for intervention, and how the output of AI intervention can help decision-making.
    harms are also covered in the CONSORT-AI list, and the team called on researchers to describe the results of an analysis of performance errors and how to identify them.
    consensus document also requires information on whether and how AI interventions and their code can be accessed, and whether there are any restrictions on access and re-use.
    Rabia Khan, who has experience with AI and data-driven drug discovery, says the CONSORT-AI statement is an industry guide to how AI and engineering teams work in the early life cycle of product development.
    , Khan said , " I believe these requirements are essential and provide the necessary framework for non-medical innovators to address medical challenges . "
    AI-specific requirements can help start-ups develop a set of requirements to meet the commercialization needs of their products, helping them clearly plan their market entry strategies.
    " Khan says that, based on her experience, different data quality and structures between different healthcare systems can pose challenges for large-scale deployment of AI algorithms.
    she said this allowed consensus statements to focus on data standards and recommendations to clearly address how data is processed within the core of the algorithm.
    , the regulation of AI algorithms should also address the diversity of data sets and the potential biases of algorithms, she said.
    one solution, suggests Khan, is to provide a comprehensive data set from regulators that allows companies to evaluate algorithms before clinical implementation.
    . Liu, X., Cruz Rivera, S., Moher, D.et al. Reporting guidelines for clinical trials for interventions involving artificial intelligence: the CONSORT-AI extension. Nat Med 26, 1364-1374 (2020).
    This article is an English version of an article which is originally in the Chinese language on echemi.com and is provided for information purposes only. This website makes no representation or warranty of any kind, either expressed or implied, as to the accuracy, completeness ownership or reliability of the article or any translations thereof. If you have any concerns or complaints relating to the article, please send an email, providing a detailed description of the concern or complaint, to service@echemi.com. A staff member will contact you within 5 working days. Once verified, infringing content will be removed immediately.

    Contact Us

    The source of this page with content of products and services is from Internet, which doesn't represent ECHEMI's opinion. If you have any queries, please write to service@echemi.com. It will be replied within 5 days.

    Moreover, if you find any instances of plagiarism from the page, please send email to service@echemi.com with relevant evidence.