Abstract: Objective: In order to explore the quantitative indicators based on CT voxel, analyze the risk factors related to the severity of chronic obstructive pulmonary disease (COPD), and establish an individualized nomogram model for predicting severe COPD. >Methods: From May 2020 to September 2021, a total of 118 eligible COPD patients were enrolled in the Affiliated Hospital of Yan’an University after dual gas phase scanning and pulmonary function tests. According to the severity classification of COPD (GOLD) (group), the patients were divided into mild COPD group (GOLDI, grade II): 66 cases; Severe COPD group (GOLD III, IV): 52 cases. The quantitative and clinical indexes of dual-gas phase registration CT between mild and severe COPD groups were compared. SPSS and R software were used for statistical analysis, and a nomogram model was established to analyze the independent risk factors of severe COPD. Results: The percentage of functional small-airway disease area (fSAD%) and the that of emphysema area (Emph%) in the severe COPD group were higher than those in the mild COPD group. With the increase of COPD severity, the pulmonary function indexes FVC, FEV1%pred, and FEV1/FVC% decreased. The smoking index, fSAD%, and Emph% of the severe COPD group were higher than those of the mild COPD group. Through analysis, it was found that the smoking index and CT quantitative indicators (fSAD%, Empire%, and Normal%) are independent risk factors for severe COPD. Furthermore, a risk model for predicting severe COPD was established. CT quantitative indicators have a relatively greater impact on the weight of patient mortality risk than smoking index. The column chart prediction model was internally validated through a calibration chart, which showed that the model calibration curve was close to the standard curve. Conclusion: Quantitative indicators based on CT voxels can predict the severity of COPD. The established column chart model for predicting COPD severity through CT quantitative indicators and smoking index has good diagnostic efficacy.