Radiomics of Vascular Structures in Pulmonary Ground-Glass Nodules: A Predictor of Invasiveness : Radiomics of Vascular Structures in GGNs for Tumor Invasiveness Prediction.
IF 1.1 4区 医学Q3 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING
{"title":"Radiomics of Vascular Structures in Pulmonary Ground-Glass Nodules: A Predictor of Invasiveness : Radiomics of Vascular Structures in GGNs for Tumor Invasiveness Prediction.","authors":"Wuling Wang, Xuan Qi, Yongsheng He, Hongkai Yang, Dong Qi, Zhen Tang, Qiong Chen","doi":"10.2174/0115734056385352250410053810","DOIUrl":null,"url":null,"abstract":"<p><p>Objective The global incidence of lung cancer highlights the need for improved assessment of nodule characteristics to enhance early detection of lung adenocarcinoma presenting as ground-glass nodules (GGNs). This study investigated the applicability of radiomics features of vascular structures within GGNs for predicting invasiveness of GGNs. Methods In total, 165 pathologically confirmed pulmonary GGNs were retrospectively analyzed. The nodules were classified into preinvasive and invasive groups and randomly categorized into training and validation sets in a 7:3 ratio. Four models were constructed and evaluated: radiomics-GGN, radiomics-vascular, clinical-radiomics-GGN, and clinical-radiomics-vascular. The predictive performance of these models was assessed using receiver operating characteristic curves, decision curve analysis, calibration curves, and DeLong's test. Results Significant differences were observed between the preinvasive and invasive groups in terms of age, nodule length, average diameter, morphology, and lobulation sign (P = 0.006, 0.038, 0.046, 0.049, and 0.002, respectively). In the radiomics-GGN model, the support vector machine (SVM) approach outperformed logistic regression (LR), achieving an area under the curve (AUC) of 0.958 in the training set and 0.763 in the validation set. Similarly, in the radiomics-vascular model, the SVM approach outperformed LR. Furthermore, the clinical-radiomics-vascular model demonstrated superior predictive performance compared with the clinical-radiomics-GGN model, with an AUC of 0.918 in the training set and 0.864 in the validation set. DeLong's test indicated significant differences in predicting the invasiveness of pulmonary nodules between the clinical-radiomics-vascular model and the clinical-radiomics-GGN model, both in the training and validation sets (P < 0.01). Conclusion The radiomics models based on internal vascular structures of GGNs outperformed those based on GGNs alone, suggesting that incorporating vascular radiomics analysis can improve the noninvasive assessment of GGN invasiveness, thereby aiding in clinical decision-making and guiding biopsy selection and treatment planning.</p>","PeriodicalId":54215,"journal":{"name":"Current Medical Imaging Reviews","volume":" ","pages":""},"PeriodicalIF":1.1000,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Current Medical Imaging Reviews","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.2174/0115734056385352250410053810","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING","Score":null,"Total":0}
引用次数: 0
Abstract
Objective The global incidence of lung cancer highlights the need for improved assessment of nodule characteristics to enhance early detection of lung adenocarcinoma presenting as ground-glass nodules (GGNs). This study investigated the applicability of radiomics features of vascular structures within GGNs for predicting invasiveness of GGNs. Methods In total, 165 pathologically confirmed pulmonary GGNs were retrospectively analyzed. The nodules were classified into preinvasive and invasive groups and randomly categorized into training and validation sets in a 7:3 ratio. Four models were constructed and evaluated: radiomics-GGN, radiomics-vascular, clinical-radiomics-GGN, and clinical-radiomics-vascular. The predictive performance of these models was assessed using receiver operating characteristic curves, decision curve analysis, calibration curves, and DeLong's test. Results Significant differences were observed between the preinvasive and invasive groups in terms of age, nodule length, average diameter, morphology, and lobulation sign (P = 0.006, 0.038, 0.046, 0.049, and 0.002, respectively). In the radiomics-GGN model, the support vector machine (SVM) approach outperformed logistic regression (LR), achieving an area under the curve (AUC) of 0.958 in the training set and 0.763 in the validation set. Similarly, in the radiomics-vascular model, the SVM approach outperformed LR. Furthermore, the clinical-radiomics-vascular model demonstrated superior predictive performance compared with the clinical-radiomics-GGN model, with an AUC of 0.918 in the training set and 0.864 in the validation set. DeLong's test indicated significant differences in predicting the invasiveness of pulmonary nodules between the clinical-radiomics-vascular model and the clinical-radiomics-GGN model, both in the training and validation sets (P < 0.01). Conclusion The radiomics models based on internal vascular structures of GGNs outperformed those based on GGNs alone, suggesting that incorporating vascular radiomics analysis can improve the noninvasive assessment of GGN invasiveness, thereby aiding in clinical decision-making and guiding biopsy selection and treatment planning.
期刊介绍:
Current Medical Imaging Reviews publishes frontier review articles, original research articles, drug clinical trial studies and guest edited thematic issues on all the latest advances on medical imaging dedicated to clinical research. All relevant areas are covered by the journal, including advances in the diagnosis, instrumentation and therapeutic applications related to all modern medical imaging techniques.
The journal is essential reading for all clinicians and researchers involved in medical imaging and diagnosis.