 |
Vol.42 No.7 contents | Japanese/English |
 | Full Text of PDF (624K) Article in Japanese |
- -
Solitary Pulmonary Nodule: Differentiating Benign From Malignant Nodule, CT-Bronchoscopic Correlation, and Usefulness of Artificial Neural Networks on CT
Hideyuki Watanabe1, Takatoshi Aoki1, Yuichi Matsuki1, Hajime Nakata11Department of Radiology, University of Occupational and Environmental Health School of Medicine, Japan.
Objective. To evaluate the useful CT finding for differentiating benign from malignant pulmonary nodule and for CT-bronchoscopic correlation, and to evaluate the usefulness of artificial neural networks (ANN). Methods. Sixty-six patients of solitary pulmonary nodule smaller than 3 cm were evaluated for differential diagnosis and CT-bronchoscopic correlation. One hundred fifty-five patients of solitary pulmonary nodule smaller than 3 cm were evaluated for ANN analysis. Results. In lung cancers, air bronchogram/bronchiologram, grossly irregular margin, and the involvement of pulmonary vein were common, while straight or inward concave margin and satellite lesions were predominantly seen in benign nodules. The positive bronchus sign on CT was useful in predicting the success of transbronchial biopsy. The average Az value for all radiologists increased with the use of the ANN output.
key words: Solitary pulmonary nodule, CT, Lung cancer, Transbronchial biopsy, Artificial neural network
JJLC 42 (7): 702-709, 2002
