[Skip to Navigation]
November 22, 2019

Lifecycle Regulation of Artificial Intelligence– and Machine Learning–Based Software Devices in Medicine

Author Affiliations
  • 1Harvard Medical School, Boston, Massachusetts
  • 2Laboratory for Technology, Markets, and Regulation, University of Zurich, Zurich, Switzerland
  • 3Division of Pharmacoepidemiology and Pharmacoeconomics, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts
  • 4Institute of Law and Laboratory for Technology, Markets, and Regulation, University of Zurich, Zurich, Switzerland
JAMA. 2019;322(23):2285-2286. doi:10.1001/jama.2019.16842

Artificial intelligence– and machine learning (AI/ML)–based technologies aim to improve patient care by uncovering new insights from the vast amount of data generated by an individual patient and by the collective experience of many patients. Machine learning is an AI technique that trains software algorithms to learn from and act on new data to continuously improve performance.1

Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    2 Comments for this article
    Software Lifecycle Regulation
    Hugh Harvey, MBBS BSc FRCR MD | Hardian Health
    The authors may wish to read up on the international standards that already exist in regards to medical device software regulation, in particular the IEC standard 62304 which specifically addresses multiple points within their article pertaining to lifecycles of software (1).

    In general, there is an ongoing trend for academics (who have never been involved in producing commercial software) penning such articles claiming that regulation either doesn’t exist or needs improving. This is due to a lack of awareness, as the academic remit does not usually extend into commercialisations and the jury-gritty of medics device software development.

    developers in industry have, for decades, been working to ISO and IEC standards, and there is an entire thriving business of quality assurance and regulatory affairs which covers these topics in immense detail.


    1. https://www.iso.org/standard/38421.html
    Response to Comment "Software Lifecycle Regulation"
    Kerstin Vokinger, MD, JD, PhD | University of Zurich, Switzerland / Division of Pharmacoepidemiology and Pharmacoeconomics, Brigham and Women’s Hospital, Harvard Medical School
    Voluntary frameworks and industry standards, including the one referenced by the commenter as well as those by other bodies, can be very useful but are not a substitute for FDA's decisions. As the FDA states in guidance to industry, the use of consensus standards "generally satisfies only a portion of a premarket submission." (1)

    The key reforms discussed in this article would require new FDA action and/or legislative action (e.g., time-limited approvals). This is consistent with the FDA's view that additional statutory authority is likely required to optimize premarket and postmarket review of AI/ML medical products. (2)

    encourage all interested stakeholders to participate in FDA's consultation on its regulatory approach to these products. (2)


    (1) FDA Guidance. https://www.fda.gov/media/71983/download
    (2) Regulations.gov Docket. https://www.regulations.gov/docket?D=FDA-2019-N-1185