The spectrum of data sharing policies in neuroimaging data repositories

Hum Brain Mapp. 2022 Jun 1;43(8):2707-2721. doi: 10.1002/hbm.25803. Epub 2022 Feb 10.


Sharing data is a scientific imperative that accelerates scientific discoveries, reinforces open science inquiry, and allows for efficient use of public investment and research resources. Considering these benefits, data sharing has been widely promoted in diverse fields and neuroscience has been no exception to this movement. For all its promise, however, the sharing of human neuroimaging data raises critical ethical and legal issues, such as data privacy. Recently, the heightened risks to data privacy posed by the rapid advances in artificial intelligence and machine learning techniques have made data sharing more challenging; the regulatory landscape around data sharing has also been evolving rapidly. Here we present an in-depth ethical and regulatory analysis that examines how neuroimaging data are currently shared against the backdrop of the relevant regulations and policies in the United States and how advanced software tools and algorithms might undermine subjects' privacy in neuroimaging data sharing. The implications of these novel technological threats to privacy in neuroimaging data sharing practices and policies will also be discussed. We then conclude with a proposal for a legal prohibition against malicious use of neuroscience data as a regulatory mechanism to address privacy risks associated with the data while maximizing the benefits of data sharing and open science practice in the field of neuroscience.

Keywords: data privacy; data re-identification; data sharing; data use agreement; neuroethics; neuroimaging.

Publication types

  • Review
  • Research Support, N.I.H., Extramural

MeSH terms

  • Artificial Intelligence*
  • Humans
  • Information Dissemination
  • Neuroimaging*
  • Policy
  • Privacy
  • United States