Problematic aspects of medical artificial intelligence. Part 1

Abstract

BACKGROUND: Artificial intelligence, like medicine, is a dynamically developing field that can be considered both a science and an art. This makes it much more difficult to use artificial intelligence compared to other technologies that come with a user manual. Research and start-ups in the field of medical artificial intelligence are rapidly multiplying: the popularity of smart mobile devices, networked applications and remote digital services is growing. However, there are still some problems that complicate the widespread use of artificial intelligence algorithms in everyday clinical practice. The reasons for this are the high cost of operating neural network platforms and the limited qualifications of some medical professionals in the field of computer technology. These are only temporary difficulties, though, which should and will be gradually resolved. CONCLUSION: This article focuses on the most sensitive points that are currently hindering the accelerated progress of machine learning in healthcare.