AUTHOR=Mun Seong K. , Wong Kenneth H. , Lo Shih-Chung B. , Li Yanni , Bayarsaikhan Shijir 

TITLE=Artificial Intelligence for the Future Radiology Diagnostic Service

JOURNAL=Frontiers in Molecular Biosciences

VOLUME=Volume 7 - 2020

YEAR=2021

URL=https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2020.614258

DOI=10.3389/fmolb.2020.614258

ISSN=2296-889X

ABSTRACT=Radiology historically has been a leader of digital transformation in healthcare. Introduction of digital imaging systems, picture archiving and communication systems (PACS) and teleradiology totally transformed radiology service over the past 30 years. Radiology is again at the crossroad for the next generation of transformation, possibly evolving as a one-stop integrated diagnostic service. Artificial intelligence and machine learning are expected to offer powerful new tools to facilitate this transformation.  The radiology community has been developing computer-aided diagnosis (CAD) tools based on machine learning (ML) technology before the advent of artificial intelligence (AI), which caught the imagination of the rest of the healthcare community.  Among various ML and AI techniques, deep-learning convolutional neural networks (CNN) and its variants have been widely used and have been proven highly effective in medical image pattern recognition.  Since the 1990s, diverse research projects have explored for CNN, and many products have been developed based on narrowly defined pattern recognition. But clinical adoption has been painfully slow due to a lack of substantial clinical relevance and sustainable business models. It is expected in the future that AI will play a more central role in radiology by offering powerful new tools to improve the overall productivity of clinical service. This paper proposes three pathways for AI's role in radiology beyond current CNN based capabilities such as, (i) improve the performance of CNN, (ii) improve the productivity of radiology service by AI-assisted workflow and (iii) develop radiomics that integrate the data from radiology, pathology, and genomics to facilitate the emergence of new integrated diagnostic service.