Explore the Application of Logistic Regression Model Based on Energy Spectral CT Parameters and Clinical Parameters Characteristics in the Diagnosis of Benign and Malignant Pulmonary Ground Glass Nodules
10.13241/j.cnki.pmb.2025.13.015
- VernacularTitle:基于能谱CT参数与临床参数特征构建的Logistic回归模型在肺磨玻璃结节良恶性诊断中的应用研究
- Author:
Fei MA
1
;
Qing-sheng SUN
;
Zhi-chuan YE
;
Huai LI
Author Information
1. 厦门市苏颂医院放射影像科 福建厦门 361000
- Publication Type:Journal Article
- Keywords:
Energy spectral CT parameters;
Clinical parameter characteristics;
Logistic regression model;
Pulmonary ground glass nodules;
Benign and malignant
- From:
Progress in Modern Biomedicine
2025;25(13):2201-2207
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To construct a Logistic regression model based on the energy spectral computed tomography(CT)parameters and clinical parameters characteristics,and to evaluate its diagnostic efficacy in the benign and malignant pulmonary ground glass nodules(GGN).Methods:162 GGN patients who were admitted to our hospital from April 2022 to March 2024 were selected,they were divided into benign group and malignant group according to pathological results.The difference of CT parameters[water content in normal scan period,water content in arterial period,slope of energy spectral curve in normal scan period and k value in arterial period]and clinical data between the two groups was compared.Multivariate Logistic regression model was used to analyze the independent influencing factors of malignant GGN.Receiver operating characteristic(ROC)curve was used to evaluate the diagnostic efficacy of the model for malignant GGN.Results:Water content in normal scan period in malignant group was significantly higher than that in benign group(P＜0.05).k value in normal scan period in the malignant group was significantly higher than that in the benign group(P＜0.05).Water content in arterial period in malignant group was higher than that in benign group(P＜0.05).k value in arterial period in the malignant group was higher than in benign group(P＜0.05).There were significant differences in smoking history,family history of tumor and chronic obstructive pulmonary disease(COPD)between benign group and malignant group(P＜0.05).Multivariate Logistic regression model analysis showed that,water content in normal scan period,k value in normal scan period,k value in arterial period,water content in arterial period,smoking history and family history of tumor were independent influencing factors of malignant nodule in GGN patients(P＜0.05).Based on the results of multi-factor Logistic regression analysis,the Logistic regression prediction model was constructed:logit(P)=ln(P/l-P)=0.015× water content in arterial period+1.214× smoking history+1.506× family history of tumor+0.013× water content in normal scan period 0.553× k value in normal scan period+0.202 × k value in arterial period.ROC curve results showed that the area under the curve(AUC)of the combined prediction model was 0.852,which was significantly higher than 0.654,0.607,0.628,0.759,0.707,0.682 of water content in normal scan period,k value in normal scan period,k value in arterial period,water content in arterial period,smoking history and family history of tumor.Conclusion:The Logistic regression model constructed based on the characteristics of energy spectral CT parameters and clinical parameters characteristics has good diagnostic efficacy for benign and malignant GGN.
