Highlights
- •As in the case of most disruptive technologies, assessment of and consensus on the possible ethical pitfalls lag.
- •selection bias in AI datasets can result in inaccurate results in under-represented patient groups.
- •AI models with imaging data acquired from one setting may poorly generalize to other practice settings.
- •Patient image data ownership regulations vary by country and domain.
- •De-identification of image data used in AI algorithms may inadvertently reveal protected health information.
Abstract
Keywords
Purchase one-time access:
Academic and PersonalCorporate R&D ProfessionalsOne-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:
Subscribe to European Journal of RadiologyReferences
Harris S, Parekh S. Funding Analysis of Companies Developing Machine Learning Solutions for Medical Imaging. Signify Research. [Internet]. Funding Analysis of Companies Developing Machine Learning Solutions for Medical Imaging. Signify Research. [cited 2019 Jun 2]. Available from: https://s3-eu-west-2.amazonaws.com/signifyresearch/app/uploads/2019/01/31125920/Funding-Analysis-of-Companies-Developing-Machine-Learning-Solutions-for-Medical-Imaging_Jan-2019.
- Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification.2018: 77-91
- AI Is the New Electricity: the Disruptive Power of AI Applications in Medical Imaging.RSNA Spotlight Course, San Francisco2019 (May 31)
- Methodologic guide for evaluating clinical performance and effect of artificial intelligence technology for medical diagnosis and prediction.Radiology. 2018; 286: 800-809
- Ownership of medical information.JAMA. 2009; 301: 1282-1284
- Health information ownership: legal theories and policy implications.Vand. J. Ent. & Tech. L. 2016; 19: 207
- Canadian Association of Radiologists white paper on artificial intelligence in radiology.Can. Assoc. Radiol. J. 2018; 69: 120-135
Geis et al. R. Ethics of AI in Radiology: the European and North American Multisociety Statement [Internet]. American College of Radiology. [cited 2019 Jun 23]. Available from: https://www.acrdsi.org/-/media/DSI/Files/PDFs/Ethics-of-AI-in-Radiology-Statement_RFC.
- GDPR Recitals and Articles [Internet].European Union, 2019 ([cited 2019 Jun 17]. Available from:)
- Privacy, confidentiality, and autonomy in psychotherapy.Neb L Rev. 1985; 64: 578
- Principles of Biomedical Ethics.Seventh, 2013
- The malicious use of artificial intelligence: Forecasting, prevention, and mitigation.arXiv preprint. 2018; (arXiv:180207228)
- Privacy and accountability in black-box medicine.Mich Telecomm & Tech. L Rev. 2016; 23: 1
Meyer D. AI Has a Big Privacy Problem And Europe’s GDPR Is About to Expose It | Fortune [Internet]. Fortune. [cited 2019 Jun 17]. Available from: https://fortune.com/2018/05/25/ai-machine-learning-privacy-gdpr/.
Roberts J. The GDPR Is in Effect: Should U.S. Companies Be Afraid? [Internet]. Yahoo FInance. [cited 2019 Jun 30]. Available from: https://finance.yahoo.com/news/gdpr-effect-u-companies-afraid-033029332.html.
- Automated de-identification of free-text medical records.BMC Med. Inform. Decis. Mak. 2008; 8: 32
- Obscuring surface anatomy in volumetric imaging data.Neuroinform. 2013; 11: 65-75
Newman L. AI Can Recognize Your Face Even If You’re Pixelated | WIRED [Internet]. Wired. [cited 2019 Jun 30]. Available from: https://www.wired.com/2016/09/machine-learning-can-identify-pixelated-faces-researchers-show/.
- Is deidentification sufficient to protect health privacy in research?.Am. J. Bioeth. 2010; 10: 3-11
- Feasibility of reidentifying individuals in large national physical activity data sets from which protected health information has been removed with use of machine learning.JAMA Network Open. 2018; 1 (e186040–e186040)
- A roadmap for foundational research on artificial intelligence in medical imaging: from the 2018 NIH/RSNA/ACR/the Academy Workshop.Radiology. 2019; 291: 781-791
