A Comparison between Inside and Outside Texture Features Extracting from Pulmonary Nodules of CT Images

Objective: To compare the differences of texture features extracting from inside and outside of pulmonary nodules of CT images. Methods: We extract texture features from 2,171 CT images of 185 cases with gray-level co-occurrence matrix and wavelet transform. Chi-square test, t test, nonparametric statistics and multiple analyses of variance are used to analyze the database. We can get basic situation of cases and differences of texture features between inside and outside of pulmonary nodules of CT images. Results: Gender distribution in cases show no statistically significant differences (X² = 2.362, P = 0.124). Texture features extracting form inside and outside of pulmonary nodules of CT images with gray-level co-occurrence matrix (F = 1642.212, P < 0.001) and wavelet transform (F = 711.572, P < 0.001) show statistically significant differences. Conclusions: There are differences in texture features between inside and outside of pulmonary nodules of CT images. We need to select the most appropriate texture features extracting region to establish prediction model.