Correlation Analysis Between CT Imaging Features and Pathological Types of Pulmonary Ground-Glass Nodules
Objective To investigate the relationship between CT features of pulmonary ground-glass nodules (GGNs) and different pathological types,providing an imaging basis for clinical treatment methods. Methods A to-tal of 165 patients with surgically and pathologically confirmed pulmonary ground-glass nodules (GGNs),with a total of 165 lesions,were retrospectively analyzed. The patients were divided into three groups based on the pathological results:Group A (benign lesions),Group B (carcinoma in situ),and Group C (including minimally invasive adenocarcinoma and in-vasive adenocarcinoma). Group A had 50 patients,Group B had 64 patients,and Group C had 51 patients. Clinical data of the three groups were collected,and the relationship between the CT signs of GGNs and pathological types was compared. Results Significant differences in CT signs such as lesion size,plain scan density,vacuole sign,spiculation,lobulation,vas-cular penetration sign,and bronchial sign were observed among the three groups (P<0.05). Specifically,significant differences in CT signs such as lesion size,plain scan density,vacuole sign,spiculation,and lobulation were observed between Group B and Group A (P<0.05). Similarly,significant differences in lesion size,plain scan density,vacuole sign,spiculation,lobula-tion,vascular penetration sign,and bronchial sign were observed between Group C and Group A (P<0.05). In the analysis of the diagnostic value of CT signs for carcinoma in situ,the AUCs for lesion size and plain scan density were 0.682 and 0.736,respectively,while the combined predictive probability of the two indicators was 0.826,which was significantly improved com-pared to each indicator alone (P<0.05). In the analysis of the diagnostic value of CT signs for minimally invasive adenocarcinoma and invasive adenocarcinoma,the AUCs for lesion size,plain scan density,and vascular penetration sign were 0.709,0.779,and 0.732,respectively,while the combined predictive probability of the two indicators was 0.868,which was significantly improved compared to each indicator alone (P<0.05). Conclusion The larger the lesion volume and the higher the density,the greater the possibility of GGN being invasive adenocarcinoma;spiculation,lobulation,and vascular penetration can differentiate the benign and malignant nature of GGNs.
Tomography,spiral computedLung adenocarcinomaGround-glass noduleLogistic modelROC curveArea under the curve
万方数据
维普
