FDA reviewed artificial intelligence-enabled products applicable to emergency medicine

被引:0
作者
Morey, Jacob [1 ]
Schupbach, John [1 ]
Jones, Derick [1 ]
Walker, Laura [1 ]
Lindor, Rachel [1 ]
Loufek, Brenna [2 ]
Mullan, Aidan [1 ]
Cabrera, Daniel [1 ]
机构
[1] Mayo Clin, Dept Emergency Med, 200 1st St SW, Rochester, MN 55905 USA
[2] Mayo Clin, Ctr Digital Hlth, Rochester, MN USA
关键词
Artificial intelligence; Machine learning; FDA; Emergency medicine; AI;
D O I
10.1016/j.ajem.2024.12.062
中图分类号
R4 [临床医学];
学科分类号
1002 ; 100602 ;
摘要
Objective: To identify and assess artificial intelligence (AI)-enabled products reviewed by the U.S. Food and Drug Administration (FDA) that are potentially applicable to emergency medicine (EM). Methods: The FDA AI-enabled products website was accessed to identify all marketed products as of March 2024. Board-certified EM physicians analyzed all products for applicability to EM practice. Inclusion criteria included products used by EM physicians directly or non-EM physicians participating directly in the evaluation and management of patients in an acute care setting. The Clinical and Economic Review (ICER) Evidence Rating Matrix was used to rate the net health benefit of applicable products. Results: A total of 882 AI-enabled products have been reviewed by the FDA from 1995 to 2024. There were 272 products that were updates of prior products that were excluded, leaving 610 unique products. Products were most commonly evaluated by Radiology (454/610), Cardiovascular (59/610), and Neurology (25/610) panels. We found 154 (25 %) products applicable to EM that were approved through Radiology (121/154), Cardiovascular (24/154), Neurology (5/154), Anesthesiology (3/154), and Ophthalmology (1/154) panels. There were 30 products that were rated as having a comparable or incremental net health benefit with moderate certainty (a C+ rating). Conclusion: An increasing number of AI-enabled products are available and regulated by the FDA. We have identified 154 that are applicable to EM, primarily related to assisting with diagnosis on various imaging modalities. There remain many opportunities for EM to assist in product reviews and meaningful translation of products into clinical practice. (c) 2024 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
引用
收藏
页码:241 / 246
页数:6
相关论文
共 16 条
  • [1] Point-of-care Ultrasonography for the Diagnosis of Acute Cardiogenic Pulmonary Edema in Patients Presenting With Acute Dyspnea: A Systematic Review and Meta-analysis
    Al Deeb, Mohammad
    Barbic, Skye
    Featherstone, Robin
    Dankoff, Jerrald
    Barbic, David
    [J]. ACADEMIC EMERGENCY MEDICINE, 2014, 21 (08) : 843 - 852
  • [2] [Anonymous], 2024, Software as a Medical Device (SaMD)
  • [3] A conceptual model of emergency department crowding
    Asplin, BR
    Magid, DJ
    Rhodes, KV
    Solberg, LI
    Lurie, N
    Camargo, CA
    [J]. ANNALS OF EMERGENCY MEDICINE, 2003, 42 (02) : 173 - 180
  • [4] Cairns C., 2021, National Hospital Ambulatory Medical Care Survey: 2018 emergency department tables
  • [5] Artificial Intelligence in Emergency Medicine: Viewpoint of Current Applications and Foreseeable Opportunities and Challenges
    Chenais, Gabrielle
    Lagarde, Emmanuel
    Gil-Jardine, Cedric
    [J]. JOURNAL OF MEDICAL INTERNET RESEARCH, 2023, 25
  • [6] fda.gov, 2024, How to Study and Market Your Device
  • [7] fda.gov, 2017, Search Registration and Listing
  • [8] How the FDA Regulates AI
    Harvey, H. Benjamin
    Gowda, Vrushab
    [J]. ACADEMIC RADIOLOGY, 2020, 27 (01) : 58 - 61
  • [9] icer.org, 2023, Value assessment framework
  • [10] FDA-Approved Artificial Intelligence and Machine Learning (AI/ML)-Enabled Medical Devices: An Updated Landscape
    Joshi, Geeta
    Jain, Aditi
    Araveeti, Shalini Reddy
    Adhikari, Sabina
    Garg, Harshit
    Bhandari, Mukund
    [J]. ELECTRONICS, 2024, 13 (03)