Abstract: Objective To explore the value of CT in the differential diagnosis of lung cancer and lung cancer combined with tuberculosis by comparing and analyzing the typical CT signs. Method The clinical and imaging data of 100 cases of lung cancer(lung cancer group) and 71 cases of lung cancer complicated with tuberculosis(lung cancer complicated with tuberculosis group) were collected from Beijing Chest Hospital, Capital Medical University retrospectively, All patients underwent plain chest CT scan and contrast -enhanced CT scan. 70 cases of simple lung cancer and 51 cases of lung cancer combined with pulmonary tuberculosis patients admitted from June 2019 to January 2023 were used as the training set, and 30 cases of simple lung cancer and 20 cases of lung cancer combined with pulmonary tuberculosis patients admitted from May 2017 to January 2019 were used as the testing set. Compare the differences in age, gender and CT imaging features between two groups of patients, and select the features with statistically significant differences. The differential diagnosis model was constructed by multi-factor logistic regression and tested on an independent test set. The model performance was evaluated by receiver operator characteristic curve (ROC curve). Result There were significant differences in age, gender, vacuole sign, vascular cluster sign, pleural effusion, pleural thickening, enhancement pattern, bronchiectasis and stenosis between the two groups (P<0.05). The differential model constructed according to these features performed well, with the area under ROC curve of 0.901 in the training set and 0.871 in the test set. Conclusion Patients with lung cancer and lung cancer combined with tuberculosis have overlapping CT signs, which makes it easy to confuse the two diseases. The established differential diagnosis model is good and helpful to distinguish the two diseases.

Key words: Lung cancer, Lung cancer with tuberculosis, Body section radiography, X-ray, Differential diagnosis