 |
Vol.61 No.4 contents | Japanese/English |
 | Full Text of PDF (1200K) Article in Japanese |
- Invited Review Article -
Current Status and Potential Utility of Artificial Intelligence in Diagnostic Imaging of Lung Cancer
Shoji Kido11Department of Artificial Intelligence Diagnostic Radiology, Osaka University Graduate School of Medicine, Japan
Computer-aided diagnoses (CADs) of lung cancer have been studied for a long time. However, the development of a highly accurate CAD for lung cancer has been difficult due to the difficulty of determining appropriate features for various opacities. With the advent of deep learning, the core technology of the third artificial intelligence (AI) boom, it has now become possible to develop CAD systems with higher accuracy and a more general purpose, and expectations concerning the utility of CADs, such as lung cancer detection and differentiation, are increasing. In addition, research on radiomics and radiogenomics, which integrate non-imaging information with imaging information, is ramping up. Many companies are actively developing CAD systems; however, they are not yet fully practical. One issue is that AI cannot explain the diagnostic process, so research on AI that can explain the reason for the diagnosis is important. Five years ago, the possibility of AI replacing radiologists was widely discussed. However, at present, the shortage of radiologists has become a problem. In order for AI to be useful in lung cancer imaging in the future, not only the technical aspects of AI development but also the response of physicians who accept it are required.
key words: Artificial intelligence, Deep learning, Machine learning, Computer-aided diagnosis, Radiomics/Radiogenomics
JJLC 61 (4): 282-288, 2021
