The incidence and mortality of lung cancer have always been high. The establishment of an effective method to predict the risk of early lung cancer is of great significance for improving the survival rate of lung cancer patients. To overcome the incompleteness of cross-sectional data, an early lung cancer risk prediction model was proposed based on synthetic analysis. 247 groups of cases from hospitals were collected for experiments, and a risk assessment model was established using synthetic analysis method combined with lung nodule benign and malignant judgment results, smoking history, family history and other lung cancer risk factors. The accuracy rate of the synthetic analysis model that did not include the imaging benign and malignant classification results was 83.20 %, and the accuracy of the synthetic analysis model that included the imaging benign and malignant classification results reached 87.40 %. Combining the results of benign and malignant lung nodules based on CT images can effectively improve the accuracy of early lung cancer risk assessment models.