Objective:To investigate the predictive value of high-resolution CT (HRCT) imaging features and Ki-67 expression levels for poorly differentiated invasive non-mucinous lung adenocarcinoma (INMA), and to construct and validate a nomogram prediction model based on these factors.Methods:A total of 210 INMA patients who underwent radical surgery at Shaanxi Provincial Cancer Hospital from July 2020 to October 2023 and obtained histopathological results were retrospectively included. Based on the degree of lesion differentiation, they were divided into well/moderately differentiated INMA group ( n=152) and poorly differentiated INMA group ( n=58). The general clinical data, HRCT imaging features and Ki-67 expression of the two groups of patients were compared. Multivariate logistic regression analysis was used to analyze independent influencing factors for poorly differentiated INMA, and a nomogram prediction model was constructed. The diagnostic efficacy of the prediction model was evaluated by the receiver operator characteristic (ROC) curve, and the prediction model was validated by consistency index (C-index) and calibration curve. Results:There were no statistically significant differences in terms of age, location, margin, lobulation, vascular convergence, and cavitation between the well/moderately differentiated INMA group and poorly differentiated INMA group (all P>0.05). There were statistically significant differences in sex ( χ2=6.65, P=0.010), Ki-67 expression ( U=2.33, P=0.021), nodule size ( t=-3.34, P=0.010), spiculation ( χ2=5.22, P=0.022), pleural indentation ( χ2=17.02, P<0.001), air bronchogram ( χ2=15.54, P<0.001) and nodule type ( χ2=59.67, P<0.001) between the two groups. Multivariate analysis showed that nodule size ( OR=1.07, 95% CI: 1.01-1.14, P=0.025), nodule type ( OR=8.23, 95% CI: 3.04-22.32, P<0.001) and Ki-67 expression ( OR=1.07, 95% CI: 1.03-1.11, P<0.001) were independent influencing factors for the occurrence of poorly differentiated INMA. A nomogram prediction model for poorly differentiated INMA was constructed based on the above indicators. ROC curve analysis showed that the area under curve of the prediction model to predict the occurrence of poorly differentiated INMA was 0.893, and the sensitivity and specificity were 89.70% and 77.60%, respectively. The C-index value of the model was 0.893. The calibration curve showed that the predicted probability was in good agreement with the actual probability. Conclusions:Nodule size, nodule type in HRCT imaging features and Ki-67 expression are independent influencing factors for the occurrence of low differentiation in INMA patients. The nomogram prediction model constructed based on the above indicators has good predictive performance.