A JOURNEY ABOUT THE ETHICS OF ARTIFICIAL INTELLIGENCE IN THE ONE GOING REVOLUTION OF HEALTHCARE SYSTEMS

Abstract

Artificial intelligence (AI) seems to be a technological keystone for setting up of new health paradigms. These paradigms are necessary to provide a high-quality health care to a growing number of people. Since 2010, the strong predictive capacities of Deep Learning have induced a new spring for AI. However, several ethical issues are required to be considered: for example, on
the one hand, on using opaque deep learning network (the black box issue), and on the other hand inherited from using statistical frameworks in medicine. This article presents on ongoing interdisciplinary research dealing with the building of ethical explanations from neural network predictions. Even if ethical explanations are provided, the current pandemic health crisis underlines the current deep learning methodology, although useful for other points, is not a panacea for helping the
understanding of an unknown virus, in particular due to the absence of historical data. Our other research activity deals on using machine learning as a mediator between medical knowledge and observations from the real world to help the understanding of new healthcare phenomena.