Comparison of computed tomography texture analysis findings of ground glass opacities detected in coronavirus disease 2019 and other viral pneumonias

Aim: To investigate the utility of computed tomography texture analysis (CTTA) findings in differentiating ground glass opacities (GGO) detected in Coronavirus Disease 2019 (COVID-19) and other viral pneumonias. Methods: 120 patients who underwent thoracic computed tomography (CT) examination for COVID-19 and non-COVID-19 viral pneumonia were included in the study. The heterogeneity of the lung parenchyma was evaluated with CT tissue analysis special software (Olea Medical, France). CTTA included features based on 8 first order intensity and 3 gray level co-occurrence matrix. Results: Contrast value was higher in COVID-19 GGO (p = 0.006). Mean, median, skewness and inverse difference moment (IDM) values were higher in non-COVID-19 GGO (p values 0, 0.001, 0.021 and 0.006, respectively) mean absolute deviation, variance and contrast values were higher in normal lung parenchyma of COVID-19 patients (p values 0.011, 0.023 and 0.006, respectively). In the ROC analysis of tissue parameters obtained from both GGO and normal lung parenchyma in both groups, the sensitivity values of tissue parameters ranged from 58.3% to 66.7%, while the specificity values ranged from 50% to 66.7%. Conclusion: Mean, median, skewness and IDM values, which are CTTA parameters, can be used to distinguish COVID-19 and non- COVID-19 viral pneumonias.