On the Interpretability of Artificial Intelligence in Radiology: Challenges and Opportunities
On the Interpretability of Artificial Intelligence in Radiology: Challenges and Opportunities
复制标题
DOI：
10.1148/ryai.2020190043
复制
发表时间：
2020-05-01
期刊：
影响因子：
--
通讯作者：
Wiest, Roland
中科院分区：
文献类型：
--
作者：
Reyes, Mauricio;Meier, Raphael;Wiest, Roland
随着人工智能（AI）系统开始进入临床放射学实践，确保它们正确运行并获得专家的信任至关重要。为了实现这一目标，使人工智能“可解释”的方法已经引起了人们的注意，以增强对机器学习算法的理解，尽管它很复杂。本文旨在深入了解放射学AI的可解释性方法的现状。这篇综述讨论了放射科医生对这一主题的看法，并提出了在临床实践中有效简化可解释性方法需要解决的趋势和挑战。本文有补充材料。(C)RSNA，2020年。
As artificial intelligence (AI) systems begin to make their way into clinical radiology practice, it is crucial to assure that they function correctly and that they gain the trust of experts. Toward this goal, approaches to make AI "interpretable" have gained attention to enhance the understanding of a machine learning algorithm, despite its complexity. This article aims to provide insights into the current state of the art of interpretability methods for radiology AI. This review discusses radiologists' opinions on the topic and suggests trends and challenges that need to be addressed to effectively streamline interpretability methods in clinical practice. Supplemental material is available for this article. (C) RSNA, 2020.