Expert Systems for Interpretable Decisions in the Clinical Domain

Abstract

This chapter presents an overview of deep learning models, highlighting some of the recent challenges with adopting AI systems, particularly deep learning, into clinical medicine. It describes the basic building blocks of computer-aided decisions, known as expert systems, and provides a perspective on how rule-based expert systems are already used in clinical practice. A roadmap is outlined for developing AI systems that benefit from mixed intelligence, incorporating both expert (domain) knowledge and the successes of deep learning. These mixed intelligence systems could promote the adaptation of more interpretable AI models into clinical medicine, overcoming the challenges faced by deep learning methods.

Ähnliche Arbeiten