Application of Gray-Scale Texture Feature in the Diagnosis of Pulmonary Nodules

Shi Chang-zheng,Zhao Qian,Luo Liang-ping
DOI: https://doi.org/10.4028/www.scientific.net/amm.140.34
2011-01-01
Applied Mechanics and Materials
Abstract:In this paper, we have proposed a new way to detect lung nodules with image texture features. 104 cases of lung nodules, including 31 benign cases and 73 malignant cases are collected, and the gray-scale correlation and texture heterogeneity are computed through CT imagings for all patients. We find that the gray correlation parameters are different between benign and malignant nodules. The heterogeneity parameters in malignant nodules are higher than that in benign noduals. The gray-scale texture correlation and heterogeneity parameters have diagnostic value in differentiating benign and malignant lung nodules. This study is an exploring study, which still needs further research.
What problem does this paper attempt to address?