Adversarial attacks on medical machine learning. | AHRQ Patient Safety Network

Adversarial attacks on medical machine learning. | AHRQ Patient Safety Network

• Study
 
• Published March 2019

Adversarial attacks on medical machine learning.

Finlayson SG, Bowers JD, Ito J, Zittrain JL, Beam AL, Kohane IS. Science. 2019;363:1287-1289.

This review article delineates how artificial intelligence approaches to medical diagnosis are vulnerable to adversarial attacks. The authors suggest that regulatory review of devices that employ machine learning include an assessment of resilience against such attacks.

PubMed citationAvailable atRelated news article

Related Resources

JOURNAL ARTICLE › STUDY

Evaluation and accurate diagnoses of pediatric diseases using artificial intelligence.

Liang H, Tsui BY, Ni H, et al. Nat Med. 2019;25:433-438.

BOOK/REPORT

Health IT Safe Practices for Closing the Loop.

Partnership for Health IT Patient Safety. Plymouth Meeting, PA: ECRI; August 2018.

MEETING/CONFERENCE › UNITED STATES MEETING/CONFERENCE

Procuring Interoperability Achieving High-Quality, Connected, and Person-Centered Care.

April 16, 2019, 3:30–4:45 PM (Eastern); National Academy of Medicine, Washington, DC.

JOURNAL ARTICLE › STUDY

Essential activities for electronic health record safety: a qualitative study.

Ash JS, Singh H, Wright A, Chase D, Sittig DF. Health Informatics J. 2019 Mar 8; [Epub ahead of print].

JOURNAL ARTICLE › COMMENTARY

A decade of health information technology usability challenges and the path forward.

Ratwani RM, Reider J, Singh H. JAMA. 2019 Feb 4; [Epub ahead of print].

View More