Objective To investigate the value of constructing nomogram model based on CT image features in the differential diagnosis of small cell lung cancer(SCLC) and non-small cell lung cancer(NSCLC).

Methods The imaging data of 120 lung cancer patients(including 39 cases of SCLC and 81 cases of NSCLC) were retrospectively analyzed, and randomly divided into the training set and validation set in the ratio of 7∶3.The univariate and multivariate logistic regression analyses were used to screen for the independent influencing factors, and a nomogram model was constructed.Propensity score matching(PSM) was applied to match 37 NSCLC patients as the control group on a 1∶1 basis, and the general clinical data and imaging characteristics between two groups were compared after PSM matching.The performance of the model was evaluated by receiver operating characteristic(ROC) curves, calibration curves, Hosmer-Lemeshow test and clinical decision curve analysis(DCA).

Results The results of univariate analysis showed that the lesion diameters >3 cm, deep lobulation signs, elongated burr signs, necrotic cavity signs, and vacuole signs in the NSCLC group were more than those in SCLC group(P < 0.05), and the level of CEA in the SCLC group was significantly higher than that in NSCLC group(P < 0.05).The results of ROC curve analysis showed that the AUC values of the training set and validation set of the model were 0.838 and 0.820, respectively, and its sensitivity and specificity were 86.50%, 80.25% and 84.37%, 79.88%, respectively.The results of Hosmer-Lemeshow test showed that the model was well fitted, and the calibration curves of the training set and validation set were close to the ideal curves, which indicated that the model had a better prediction accuracy.The DCA analysis of the training set and validation set indicated that the nomogram model had a good net benefit in discriminating SCLC from NSCLC in clinical applications.After PSM matching, the results of univariate analysis showed that the percentage of lesions >3 cm in diameter, necrotic cavities, and deep lobular signs in the NSCLC group were more than those in SCLC group(P < 0.05).

Conclusions The construction of nomogram model based on CT image features has certain accuracy, calibration and differentiation in identifying SCLC and NSCLC, which provides a reference basis for clinical development of targeted personalized treatment plans.