Emerging vulnerabilities demand new conversations With public and academic attention increasingly focused on the new role of machine learning in the health information economy, an unusual and no-longer-esoteric category of vulnerabilities in machine-learning systems could prove important. These vulnerabilities allow a small, carefully designed change in how inputs are presented to a system to completely alter its output, causing it to confidently arrive at manifestly wrong conclusions. These advanced techniques to subvert otherwise-reliable machine-learning systems—so-called adversarial attacks—have, to date, been of interest primarily to computer science researchers (1). However, the landscape of often-competing interests within health care, and billions of dollars at stake in systems' outputs, implies considerable problems. We outline motivations that various players in the health care system may have to use adversarial attacks and begin a discussion of what to do about them. Far from discouraging continued innovation with medical machine learning, we call for active engagement of medical, technical, legal, and ethical experts in pursuit of efficient, broadly available, and effective health care that machine learning will enable.
[1]
Lucy Rosenbloom.
arXiv
,
2019,
The Charleston Advisor.
[2]
R. Dutton,et al.
Comparison of Anesthesia Times and Billing Patterns by Anesthesia Practitioners
,
2018,
JAMA network open.
[3]
Fabio Roli,et al.
Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning
,
2017,
Pattern Recognit..
[4]
Anne E Carpenter,et al.
Opportunities and obstacles for deep learning in biology and medicine
,
2017,
bioRxiv.
[5]
M. Wendy Hennequin,et al.
The Future of the Internet and How to Stop It
,
2011
.
[6]
Kristi Reynolds,et al.
Metabolic syndrome: underrated or underdiagnosed?
,
2005,
Diabetes care.
[7]
Aaron S Kesselheim,et al.
Overbilling vs. downcoding--the battle between physicians and insurers.
,
2005,
The New England journal of medicine.
[8]
I. Wilson,et al.
Physician manipulation of reimbursement rules for patients: between a rock and a hard place.
,
2000,
JAMA.
[9]
D. Clark,et al.
End-to-end arguments in system design
,
1984,
TOCS.
[10]
Jonathan Zittrain.
The Future of the Internet - And How to Stop It
,
2008
.