 |
Vol.59 No.1 contents | Japanese/English |
 | Full Text of PDF (1151K) Article in Japanese |
- Invited Review Article -
Quantitative Computed Tomography Imaging of Lung Cancer
Masahiro Yanagawa1, Noriyuki Tomiyama11Department of Diagnostic and Interventional Radiology, Osaka University Graduate School of Medicine, Japan
Morphological evaluations by radiologists are essential for the diagnosis of lung cancer. Evaluations of the nodule margin, internal characteristics, and the relationship with the pre-existing lung enable the imaging diagnosis of lung cancer. However, in the field of imaging diagnoses, quantification has long been emphasized. At the 2007 meeting of the Radiological Society of North America, the Quantitative Imaging Biomarkers Alliance (QIBA) was established as a cooperative standardizing organization for biomarkers in quantitative images. The QIBA aims to create standardized guidelines that can be used to objectively measure and evaluate images or parameters in clinical trials and daily clinical practice. In a clinical setting, we not only measure the dimension of nodules but also perform various quantifications, such as volumetry. While the usefulness of many quantitative indicators has been reported, one- and two-dimensional analyses are highly versatile but subjective in their utility. Three-dimensional analyses are indispensable for more objective and reproducible analyses. In recent years, studies using artificial intelligence have been performed increasingly frequently. In this paper, focusing on the malignancy and prognosis of lung cancer, we review some analyses using quantitative methods and artificial intelligence in lung cancer.
key words: Lung cancer, Computed tomography, Quantification, Artificial intelligence
JJLC 59 (1): 29-36, 2019
