like training artificial intelligence algorithms, will require
developers to protect patients’ privacy.
Clinical data should be treated as a public good when used for
research or artificial intelligence algorithm development, so long
as patients’ privacy is protected, according to a report from
the Radiological Society of North America (RSNA).
As artificial intelligence and machine learning are increasingly
applied to medical imaging, bringing the potential for
streamlined analysis and faster diagnoses, the industry still lacks
a broad consensus on an ethical framework for sharing this
data.
“Now that we have electronic access to clinical data and the
data processing tools, we can dramatically accelerate our ability
to gain understanding and develop new applications that can benefit
patients and populations,” said study lead author David
B. Larson, MD, MBA, from the Stanford University School of
Medicine. “But unsettled questions regarding the ethical use of
the data often preclude the sharing of that information.”
To offer solutions around data sharing for AI
development, RSNA developed a framework that highlights how to
ethically use patient data for secondary purposes.
“Medical data, which are simply recorded observations, are
acquired for the purposes of providing patient care,” Larson
said.
Jessica Kent, March 30, 2020
Health Analytics