J Korean Soc Radiol.
2024 Sep;85(5):834-847. 10.3348/jksr.2024.0118.
Explainable & Safe Artificial Intelligence in Radiology
- Do S
1,2,3
- Affiliations
-
- 1Laboratory of Medical Imaging and Computation, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA
- 2KU-KIST Graduate School of Converging Science and Technology at Korea University, Seoul, Korea
- 3Kempner Institute, Harvard University, Boston, MA, USA
- KMID: 2559969
- DOI: http://doi.org/10.3348/jksr.2024.0118
Abstract
- Artificial intelligence (AI) is transforming radiology with improved diagnostic accuracy and efficiency, but prediction uncertainty remains a critical challenge. This review examines key sources of uncertainty—out-of-distribution, aleatoric, and model uncertainties—and highlights the importance of independent confidence metrics and explainable AI for safe integration. Independent confidence metrics assess the reliability of AI predictions, while explainable AI provides transparency, enhancing collaboration between AI and radiologists. The development of zero-error tolerance models, designed to minimize errors, sets new standards for safety. Addressing these challenges will enable AI to become a trusted partner in radiology, advancing care standards and patient outcomes.
Keyword
- Full Text Links
-
- Actions
-
Cited
- CITED
-
- Close
- Share
-
- Similar articles
-
- Preface for Special Issue on Explainable/Reliable Artificial Intelligence, and Generative Artificial Intelligence with Large Language Model for Radiologist
- Explainable artificial intelligence in emergency medicine: an overview
- Artificial Intelligence in Medicine: Beginner's Guide
- The Impact of Artificial Intelligence on Radiology Education in the Wake of Coronavirus Disease 2019
- Role of Artificial Intelligence in Achieving Universal Health Coverage: A Mongolian Perspective
