Research on the Prediction of the Pathological Grade of Invasive Lung Adenocarcinoma by the CT Signs Model of Pulmonary Nodules
10.11969/j.issn.1673-548X.2025.06.014
- VernacularTitle:肺结节CT征象模型预测浸润性腺癌病理分级研究
- Author:
Zijun MEI
1
;
Kai JI
;
Junyan YUE
Author Information
1. 453100 新乡医学院第一附属医院放射科;454001 焦作,河南理工大学第一附属医院CT室
- Publication Type:Journal Article
- Keywords:
Invasive lung adenocarcinoma;
CT signs;
Pathological grade
- From:
Journal of Medical Research
2025;54(6):76-81
- CountryChina
- Language:Chinese
-
Abstract:
Objective A binary Logistic regression model was developed to forecast the pathological grade of invasive adenocarcino-ma by utilizing the CT characteristics of lung nodules.Methods A retrospective analysis was conducted on the clinical data,pathological types,and imaging findings of 303 cases of ground-glass nodules diagnosed with postoperative pathological infiltrative adenocarcinoma at the First Affiliated Hospital of Henan Polytechnic University and the First Affiliated Hospital of Xinxiang Medical College from January 2021 to February 2023.Based on the pathological results,these lesions were categorized into two groups:the low-grade group(compri-sing 262 cases characterized by adherent,acinar,or papillary types as predominant forms of adenocarcinoma with no more than 20％high-grade pattern)and the high-grade group(consisting of 41 cases exhibiting any form of adenocarcinoma with over 20％high-grade components).The Mann-Whitney U test was employed to compare quantitative parameters between both groups,while qualitative parameters were analyzed using the x2 test.Additionally,binary Logistic regression models were utilized to identify independent predictors;further evaluation included area under curve(AUC)values,calibration curves,and decision analysis curves to assess model differentia-tion,calibration accuracy,and clinical applicability.Results Univariate analysis revealed that gender,air bronchial sign,vacuole sign,vascular cluster sign,pleural depression sign,long diameter,short diameter,and CT-enhanced net increment exhibited statistical signifi-cance(P＜0.05),whereas location,burr sign,and solid component ratio did not demonstrate statistical significance(P＞0.05).Binary Logistic regression analysis identified long diameter,CT-enhanced net increment,vascular cluster sign,pleural depression sign,and vacu-ole sign as independent predictors of the pathological grade model for invasive adenocarcinoma.The results of ROC curve analysis indicated that the AUC value of the Logistic regression model was 0.846 with a sensitivity of 81.25％and specificity of 86.52％.Conclusion The logistic regression model based on CT signs has excellent ability and stability in predicting the pathological grade of invasive adenocarcinoma.
