Advance of CT Texture Feature Analysis in Diagnosis of Solitary Pulmonary Nodules

Lung cancer is the most common malignant tumor in China. Chest CT scan can improve the detection and correct diagnosis of lung cancer. Early or asymptomatic lung cancer is mostly manifested as solitary pulmonary nodules, so the differential diagnosis of pulmonary nodules is very important. The diagnosis of pulmonary nodules on conventional CT images depended on their contour, shape and the characteristics of density, which is difficult to differential diagnosis in benign or malignant nodules, especial in ground glass nodule (GGN). However, computed tomography (CT) texture feature analysis is based on the quantification of internal gray scale features and other features, which provides a useful reference for the identification of benign and malignant solitary pulmonary nodules, prognosis judgment and gene mutation prediction. Thus texture analysis makes up for the deficiency of traditional CT quantitative evaluation. This paper summarizes the basic principle, method and workflow of CT texture feature analysis and its application in the diagnosis of solitary pulmonary nodules.