Risk prediction model for postoperative brain metastasis in IIB-IIIB non-small cell lung cancer: based on radiomics and clinicopathology

Risk prediction model for postoperative brain metastasis in IIB-IIIB non-small cell lung cancer: based on radiomics and clinicopathology | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Risk prediction model for postoperative brain metastasis in IIB-IIIB non-small cell lung cancer: based on radiomics and clinicopathology Lu yang, Chengde Liao, Zhiqiang Ouyang, Qinqing Li, Yunchao Huang, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3972347/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Purpose To develop and validate a model based on radiomics and clinicopathological features for predicting postoperative brain metastasis (BM) in stage IIB-IIIB non-small cell lung cancer (NSCLC) patients. Materials and methods A total of 333 NSCLC patients operated from October 2015 and December 2019 with postoperative pathological stage IIB-IIIB were included, which were randomly divided into a training and validation cohort. The intratumoral and peritumoral radiomics features from preoperative CT image were extracted and selected using the least absolute shrinkage and selection operator (LASSO). The independent clinical predictors of BM were identified by univariate and multivariate Cox analysis. The radiomics model, clinical model and radiomics combined clinicopathological model were constructed with six different algorithms. Subsequently, we constructed a dynamic nomogram. The performance of the model was evaluated by the area under the curve (AUC), sensitivity, specificity, calibration curve and decision curve analysis (DCA). Results The radiomics model combining intratumoral and peritumoral radiomics features exhibited great predictive performance for BM prediction, with an AUC of 0.888–0.928 in the training cohort and 0.838–0.894 in the validation cohort. The model including the intra- and peritumoral radiomics, T stage, histological type, spiculation and other metastatic sites yielded AUC of 0.947–0.979 in the training cohort and 0.847–0.926 in the validation cohort, with good calibration for all algorithms (p > 0.05). DCA revealed that the combined model obtained a greater net benefit. Conclusion The model that integrates radiomics features with clinicopathological features could aid in early-stage prediction of postoperative BM risk in stage IIB-IIIB NSCLC patients. Dynamic nomogram provides great convenience for clinicians to manage patients. Radiomics Non-small cell lung cancer Brain metastasis Machine learning Dynamic nomogram Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Brain metastases (BM) are the predominant distant metastases in non-small cell lung cancer (NSCLC)(Achrol et al. 2019 ). Once a patient develops BM, the median progression-free survival is typically only 1–2 months, and the one-year survival rate can reach as low as 10% − 20%(Smith et al. 2019 ), which poses a significant threat to patient survival and adversely affects quality of life(Ouyang et al. 2020 ; Witlox et al. 2018 ). Early prevention and diagnosis are crucial for improving patient outcomes. Prophylactic whole-brain irradiation has been proven effective at reducing the incidence of BM (Gore et al. 2011 ; Witlox et al. 2018 ), but it may also lead to certain adverse reactions (Chalubinska et al. 2021). The challenge for clinicians lies in accurately weighing the benefits and drawbacks of radiotherapy while identifying high-risk patients who would benefit from early intervention(Chalubinska et al. 2021; Hou et al. 2022 ). Therefore, precise prediction and stratification of BM can assist clinicians in implementing targeted preventive measures. Radiomics, as a noninvasive method, enables the capture of features in CT images that are imperceptible to human eyes and quantitatively evaluate key information about the entire tumor, which can correlate these features with the prognosis of patients with cancer(Lambin et al. 2012 ; Lubner et al. 2017 ). At present, many radiomics-based models have emerged to predict the prognosis of patients, including BM(An et al. 2018; F. Sun et al. 2021 ). However, the predictive efficiency of the model needs to be further improved, so it is urgent to further explore the factors to improve the predictive efficiency of the model. At present, some researchers have proposed that tumors can invade the adjacent lung parenchyma when they spread through vascular, lymphatic, or airway infiltration(Dercle et al. 2020 ). Conversely, lung tissue around the tumor may change with the migration of tumor cells. Therefore, the evaluation of the lung tissue surrounding the tumor may have some significance for the occurrence of BM. However, these microscopic features are difficult to recognize with the naked eye. Therefore, the use of radiomics to extract and quantify the characteristics of the tissue around the tumor can help us analyze the value of the peritumor tissue. At present, some researchers have found that combined intra- and peritumoral radiomics features have demonstrated the ability to predict lymphovascular infiltration and overall survival in patients with NSCLC(Chen et al. 2022 ). The peritumour radiomics features may offer more valuable information than visible tumor features for risk stratification of distant metastases in locally advanced NSCLC patients(Lee et al. 2018). Therefore, developing a radiomic model that explores peritumour features holds significant importance for precision oncology. Radiomics can use different machine learning algorithms to build models, but the main reason that affects the use of models is the complexity of these models themselves, and their operation process is difficult to explain, which brings great challenges to model users. Currently, several interpretability methods in machine learning can be used to understand and interpret the predictions made by models, such as partial dependence graphs (PDP)(Abbas et al. 2023) and SHapley Additive exPlanations (SHAP)(Mitchell et al. 2022 ). One disadvantage of PDP is that struggle to illustrate relationships between multiple features and predictions(Angelini et al. 2023 ). SHAP values, rooted in game theory’s Shapley value concept, fairly measure features’ contributions to predictions in machine learning(Gebreyesus et al. 2023 ). For each prediction sample, the model produces a predicted value. In addition, each feature of this sample will have a corresponding SHAP value, which indicates how the feature contributes to the prediction of the specified data point. Some factors will have a positive impact on the forecast probability, while others will have a negative impact on it. Calculating SHAP values improves interpretability and helps clinicians understand the output predicted by the model. Our study aims to develop and validate intratumoral and peritumoral radiomics models, clinical models, and combined models predicting postoperative BM risk in NSCLC patients. Additionally, to better understand how radiomics features distinguish BM, we calculated the SHAP values for each feature. Our ultimate goal is to provide clinicians with an accurate probability of BM risk so that they can identify high-risk patients for early intervention. Materials and methods Patient The retrospective analysis included 4803 NSCLC patients who underwent lung cancer resection at Yunnan Cancer Hospital between October 2015 and December 2019. The inclusion criteria for patients were as follows: 1) had NSCLC whose postoperative pathological stage was IIB-IIIB, 2) underwent chest CT examination at our hospital before surgery, 3) underwent brain MRI examination before surgery to exclude BM, and 4) underwent regular CT and brain MRI examination after surgery. The exclusion criteria for patients were as follows: 1) had other malignant tumors, 2) received neoadjuvant therapy, 3) had atelectasis and difficultly determining tumor boundaries, and 4) had less than 3 years of follow-up except for patients with BM. Finally, 333 patients with stage IIB-IIIB NSCLC were included, 97 with BM and 236 without BM. The patients were randomly divided into training (n = 234) and validation (n = 99) cohorts at a 7:3 ratio which was shown in Supplemental Fig. S1 . This study was approved by the Ethics Committee of Yunnan Cancer Hospital, and the need for informed consent was waived. Follow-up Patients underwent craniocerebral MRI every 3–6 months during the initial two postoperative years, followed by annual or symptom-driven assessments, with a minimum 3-year follow-up. We chose 3 years as the cutoff point for BM follow-up because previous studies have shown that the cumulative incidence of BM in NSCLC patients without baseline BM reaches a stable level 3 years after diagnosis(Ji et al. 2014 ; Zhou et al. 2020 ), indicating that most NSCLC BM develop within 3 years of follow-up. The follow-up ended in December 2022. The endpoints of the study were considered to be BM, which was considered to occur when 1) MRI manifestations of BM were observed by two radiologists, 2) the patient presented obvious clinical symptoms, and the imaging findings of BM were observed by the radiologist; 3) the pathological diagnosis after intracranial space occupation was confirmed as BM; and 4) when the lesion was suspected to be BM increased or decreased in volume after treatment during MRI follow-up, the patient was diagnosed with BM. Image acquisition and CT semantic feature extraction The chest CT was examined using a Siemens 128 scanner with the following acquisition parameters: layer thickness, 1.0 mm; tube voltage, 120 KV; and tube current, 160 mA. All of the images were reconstructed using a sharp core (b60) and a standard lung window (window width: 1600 HU; window level: -700 HU). Two radiologists with 8 and 3 years of experience independently evaluated the CT images. The CT semantic features included the tumor location, maximum tumor diameter, presence or absence of ground glass density, air bronchial signs, lobulation, spiculation, pleural indentation, vacuoles, cavities, emphysema, pleural effusion, and other metastatic sites (see Supplemental S-1 for specific definitions). CT Image segmentation and feature extraction and selection The procedure of this study is shown in Fig. 1 . Two radiologists with 8 and 3 years of experience manually segmented the region of interest (ROI) and extracted radiomics features in 3D-Slicer (version 5.1.0. https://www.slicer.org ). The peritumoral region of interest (ROI) extended outward by 3 mm along the edge of the primary tumor, and the outward expansion distances, including the ribs, chest wall muscle and soft tissue, were manually removed. The main feature categories extracted included shape features, first-order features, gray-level co-occurrence matrix (GLCM), gray-level run-length matrix (GLRLM), gray-level size zone matrix (GLSZM), neighboring gray tone difference matrix (NGTDM) and wavelet-based features. The intraclass correlation coefficient (ICC) was calculated to evaluate the stability and repeatability of feature extraction, and features with an inter-observer ICC greater than 0.8 were included in the study. The included radiomics features were standardized. First, the training cohort data were standardized, and the standardized parameters were applied to the validation cohort. The standardized value Z for each patient-specific characteristic value X was calculated by the formula Z = X- mean/standard deviation. The mean and standard deviation were generated from the data in the training cohort. To avoid multicollinearity, correlation analysis was performed on radiomics features, and features with an absolute Pearson correlation coefficient greater than 0.85 were retained. The Least absolute shrinkage and selection operator (LASSO) was used to select the retained features, and the 10-fold cross-validation method was used to verify and select the best λ value. Clinicopathological features The clinicopathological factors analyzed included sex, smoking status, age at operation, mode of operation, pathological stage, pathological type, carcinoembryonic antigen (CEA) level, postoperative treatment, and postoperative metastasis. In the training cohort, univariate Cox analysis was used to analyze the differences between various factors in the presence and absence of BM, and variables with p < 0.05 were included in multivariate Cox regression analysis. Model construction and performance evaluation The radiomics, clinicopathological and CT semantic features were ultimately selected to construct 5 models. The models were as follows: radiomics model 1 (intratumoral radiomics features), radiomics model 2 (3 mm peritumoral radiomics features), radiomics model 3 (intratumoral combined with peritumoral radiomics features), clinical model (clinicopathological features and CT semantic features), and combined model (radiomics features, clinicopathological features and CT semantic features). A linear combination of the screened features and their weight coefficients was used to construct a Rad-score for each patient. In addition, four commonly used machine learning algorithms, namely, support vector machine (SVM), generalized linear model (GLM), linear discriminant analysis (LDA) and quadratic discriminant analysis (QDA), are used to filter the best classifier. A nomogram and SHAP value graph were constructed. The performance of the model was evaluated by the area under the curve (AUC), sensitivity, specificity, calibration curve, and decision curve analysis (DCA), and the model was validated using validation cohorts. Statistical analysis Statistical analysis was performed using R software (version 4.3.1; https://www.r-project.org ). The continuous variables subject to a normal distribution are expressed as the mean ± standard deviation, while those not subject to a normal distribution are expressed as the median and interquartile range (IQR). Chi-square tests were used to compare differences between categorical variables. The independent sample t test was used for continuous variables with a normal distribution. The clinicopathological features were screened by univariate and multivariate Cox regression analyses. The area under the curve (AUC), sensitivity and specificity were calculated to quantify the performance of the model. The AUCs were compared using the Delong test. A calibration curve was drawn to evaluate the goodness of fit of the model, and the Hosmer–Lemeshow test was used to evaluate the consistency of the predicted probability with the actual observed value. Finally, DCA was performed to evaluate the clinical utility of the prediction model by quantifying the net benefit when considering different threshold probability capabilities. The risk stratification ability of the nomograms was evaluated using Kaplan‒Meier (KM) survival curves and log-rank tests. p < 0.05 was considered to indicate statistical significance. Results Clinical characteristics of patients A total of 333 patients with NSCLC were enrolled in this study (October 2015 to December 2019). We randomly divided 333 patients into a training cohort (n = 234) and a validation cohort (n = 99). The clinicopathological features and CT semantic features of the patients are shown in Table 1 . Among the 333 patients, 202 (60.7%) were males and 131 (39.3%) were females. There were 97 patients with BM and 236 patients without BM. Across the cohort, the median follow-up time was 3.32 years (2.24, 4.49, IQR), and the median time to onset of BM was 1 year (0.63, 1.87, IQR). Moreover, there was no significant difference in the distribution of each feature between the training cohort and the validation cohort. Table 1 Demographic and clinical pathological characteristics Characteristics Training cohort (N = 234) Validation cohort (N = 99) Total (N = 333) P value Age (years) 55.2(± 8.73) 56.0(± 9.48) 55.4(± 8.96) 0.453 Sex 0.144 male 136 (58.1%) 66 (66.7%) 202 (60.7%) female 98 (41.9%) 33 (33.3%) 131 (39.3%) Smoking 0.805 no 124 (53.0%) 51 (51.5%) 175 (52.6%) yes 110 (47.0%) 48 (48.5%) 158 (47.4%) Surgery way 0.793 thoracotomy 98 (41.9%) 43 (43.4%) 141 (42.3%) thoracoscopy 136 (58.1%) 56 (56.6%) 192 (57.7%) Pathological stage 0.337 IIB 87 (37.2%) 37 (37.4%) 124 (37.2%) IIIA 118 (50.4%) 55 (55.6%) 173 (52.0%) IIIB 29 (12.4%) 7 (7.1%) 36 (10.8%) T 0.208 1 31 (13.2%) 19 (19.2%) 50 (15.0%) 2 82 (35.0%) 31 (31.3%) 113 (33.9%) 3 87 (37.2%) 29 (29.3%) 116 (34.8%) 4 34 (14.5%) 20 (20.2%) 54 (16.2%) N 0.429 0 82 (35.0%) 34 (34.3%) 116 (34.8%) 1 43 (18.4%) 25 (25.3%) 68 (20.4%) 2 103 (44.0%) 39 (39.4%) 142 (42.6%) 3 6 (2.6%) 1 (1.0%) 7 (2.1%) Histological type 0.909 squamous carcinoma 40 (17.1%) 15 (15.2%) 55 (16.5%) adenocarcinoma 187 (79.9%) 81 (81.8%) 268 (80.5%) other 7 (3.0%) 3 (3.0%) 10 (3.0%) Tumor location 0.228 superior lobe of right lung 61 (26.1%) 34 (34.3%) 95 (28.5%) middle lobe of right lung 14 (6.0%) 5 (5.1%) 19 (5.7%) inferior lobe of right lung 53 (22.6%) 13 (13.1%) 66 (19.8%) superior lobe of left lung 70 (29.9%) 34 (34.3%) 104 (31.2%) inferior lobe of left lung 36 (15.4%) 13 (13.1%) 49 (14.7%) Maximum tumor diameter (cm) 4.53(± 1.86) 4.16(± 2.00) 4.42(± 1.91) 0.108 GGO 0.894 no 202 (86.3%) 86 (86.9%) 288 (86.5%) yes 32 (13.7%) 13 (13.1%) 45 (13.5%) Air bronchogram 0.82 no 166 (70.9%) 69 (69.7%) 235 (70.6%) yes 68 (29.1%) 30 (30.3%) 98 (29.4%) Lobulated 0.511 no 27 (11.5%) 9 (9.1%) 36 (10.8%) yes 207 (88.5%) 90 (90.9%) 297 (89.2%) Spiculation 0.053 no 82 (35.0%) 24 (24.2%) 106 (31.8%) yes 152 (65.0%) 75 (75.8%) 227 (68.2%) Pleural indention 0.697 no 76 (32.5%) 30 (30.3%) 106 (31.8%) yes 158 (67.5%) 69 (69.7%) 227 (68.2%) Cavity 0.404 no 208 (88.9%) 91 (91.9%) 299 (89.8%) yes 26 (11.1%) 8 (8.1%) 34 (10.2%) Vacuole 0.811 no 184 (78.6%) 79 (79.8%) 263 (79.0%) yes 50 (21.4%) 20 (20.2%) 70 (21.0%) Emphysema 0.545 no 213 (91.0%) 88 (88.9%) 301 (90.4%) yes 21 (9.0%) 11 (11.1%) 32 (9.6%) Pleural effusion 0.366 no 228 (97.4%) 98 (99.0%) 326 (97.9%) yes 6 (2.6%) 1 (1.0%) 7 (2.1%) Other metastatic sites 0.353 no 199 (85.0%) 88 (88.9%) 287 (86.2%) yes 35 (15.0%) 11 (11.1%) 46 (13.8%) CEA 0.177 normal 94 (40.2%) 32 (32.3%) 126 (37.8%) rise 140 (59.8%) 67 (67.7%) 207 (62.2%) Fellow time (days) 1196.5(803–1596) 1232(909-1675.5) 1212(818–1627) 0.378 GGO, ground-glass opacity; CEA, carcinoembryonic antigen Radiomics feature selection and radiomics model efficiency A total of 1316 features were extracted from the intratumoral ROI, and 1302 features were extracted from the ROI 3 mm per peritumor. A total of 601 ineligible features (those with an ICC less than 0.8) were excluded to ensure the stability and repeatability of the radiomics features during segmentation. To address collinearity among radiomics features, Pearson correlation analysis was used to determine that 1100 radiomics features were significantly correlated (r > 0.85), and one of the features was retained. LASSO was used to select the best features, and the optimal λ values selected for intratumoral, peritumoral, and intratumoral combined peritumoral radiomics features were 0.02458, 0.02273, and 0.04359, respectively. Finally, the optimal λ values produced 15, 10, and 11 features, respectively, which were used to construct the radiomics model (Supplemental Fig.S2). The Rad-score calculation formula is reported in Supplemental S-2. Three radiomics models were constructed using four machine learning classifiers. In addition, a rad-score based on radiomics features was constructed and validated. With respect to radiomics model 1, the model efficiency of the QDA machine learning algorithm was greater than that of the other algorithms in the training cohort; the AUC was 0.833 (95% CI, 0.779–0.886), and the AUC was 0.646 (95% CI, 0.538–0.754) in the validation cohort (Fig. 2 ). With respect to radiomics model 2, the SVM and GLM algorithms had higher model efficiency than the other algorithms in the training cohort. The efficiency of radiomics model 3 was greater than that of radiomics models 1 and 2. With respect to the training cohort, the model based on the QDA machine learning algorithm had greater efficacy, with an AUC of 0.928 (95% CI, 0.891–0.964) in the training cohort, the AUC was 0.838 (95% CI, 0.731–0.945) in the validation cohort (Table 3 ). Table 3 Model performance comparison Training cohort Validation cohort model Accuracy Sensitivity Specificity AUC (95%CI) Accuracy Sensitivity Specificity AUC (95%CI) SVM Radiomics model 1 0.782 0.586 0.866 0.779 (0.714 to 0.843) 0.727 0.741 0.722 0.736 (0.622 to 0.84) Radiomics model 2 0.838 0.857 0.829 0.903 (0.859 to 0.946) 0.798 0.778 0.806 0.832 (0.742 to 0.92) Radiomics model 3 0.915 0.800 0.963 0.903 (0.854 to 0.952) 0.838 0.741 0.875 0.864 (0.785 to 0.94) Clinical model 0.684 0.714 0.671 0.719 (0.646 to 0.792) 0.616 0.852 0.528 0.705 (0.599 to 0.81) Combined model 0.923 0.829 0.963 0.958 (0.934 to 0.983) 0.899 0.778 0.944 0.926 (0.870 to 0.98) GLM Radiomics model 1 0.778 0.600 0.854 0.788 (0.724 to 0.852) 0.747 0.704 0.764 0.736 (0.624 to 0.84) Radiomics model 2 0.829 0.857 0.817 0.903 (0.861 to 0.944) 0.778 0.778 0.778 0.833 (0.740 to 0.92) Radiomics model 3 0.880 0.814 0.909 0.916 (0.872 to 0.961) 0.859 0.778 0.889 0.882 (0.808 to 0.95) Clinical model 0.684 0.671 0.689 0.718 (0.647 to 0.790) 0.616 0.852 0.528 0.706 (0.603 to 0.80) Combined model 0.919 0.857 0.945 0.965 (0.944 to 0.986) 0.899 0.778 0.944 0.918 (0.857 to 0.97) LDA Radiomics model 1 0.739 0.671 0.768 0.777 (0.712 to 0.841) 0.657 0.815 0.597 0.695 (0.583 to 0.80) Radiomics model 2 0.786 0.871 0.750 0.897 (0.855 to 0.939) 0.788 0.815 0.778 0.830 (0.737 to 0.92) Radiomics model 3 0.863 0.800 0.890 0.905 (0.860 to 0.950) 0.828 0.815 0.833 0.894 (0.827 to 0.96) Clinical model 0.684 0.671 0.689 0.718 (0.646 to 0.789) 0.616 0.852 0.528 0.706 (0.603 to 0.80) Combined model 0.906 0.886 0.915 0.956 (0.931 to 0.982) 0.869 0.815 0.889 0.923 (0.868 to 0.97) QDA Radiomics model 1 0.731 0.871 0.671 0.833 (0.779 to 0.886) 0.576 0.926 0.444 0.646 (0.538 to 0.75) Radiomics model 2 0.838 0.714 0.890 0.898 (0.856 to 0.940) 0.798 0.593 0.875 0.734 (0.606 to 0.86) Radiomics model 3 0.872 0.814 0.896 0.928 (0.891 to 0.964) 0.848 0.778 0.875 0.838 (0.731 to 0.94) Clinical model 0.722 0.657 0.750 0.763 (0.699 to 0.828) 0.576 0.852 0.472 0.695 (0.588 to 0.80) Combined model 0.927 0.943 0.921 0.979 (0.964 to 0.995) 0.838 0.741 0.875 0.847 (0.743 to 0.95) Radscore Radiomics model 1 0.667 0.786 0.616 0.761 (0.696 to 0.826) 0.616 0.926 0.500 0.735 (0.633 to 0.83) Radiomics model 2 0.829 0.757 0.860 0.889 (0.844 to 0.935) 0.828 0.815 0.833 0.862 (0.779 to 0.94) Radiomics model 3 0.846 0.786 0.872 0.888 (0.842 to 0.935) 0.838 0.852 0.833 0.886 (0.813 to 0.95) Clinical COX Clinical model 0.756 0.614 0.817 0.783 (0.719 to 0.847) 0.657 0.593 0.681 0.657 (0.540 to 0.77) Radscore clinical COX Combined model 0.880 0.914 0.866 0.951 (0.926 to 0.977) 0.879 0.852 0.889 0.905 (0.834 to 0.97) AUC, area under the curve; CI, confidence interval; SVM, support vector machine; GLM, generalize linear model; LDA, linear discriminant analysis; QDA, quadratic discriminant analysis Selection of clinical features and efficacy of clinical models Univariate and multivariate Cox regression analyses were used to evaluate the significance of these features in predicting BM. The results showed that T stage, pathological type, spiculation, and the presence of other metastatic sites before BM were independent predictors of BM ( p < 0.05) (Table 2 ). After these features were incorporated into the clinical model, the Cox proportional hazard regression model showed greater efficacy in the training cohort than in the validation cohort, with an AUC of 0.783 (95% CI, 0.719–0.847) and an AUC of 0.657 (95% CI, 0.540–0.774) (Table 3 ). Table 2 Uni- and multivariable cox regression analyses in training cohort comprising IIB-IIIB non–small cell lung cancer patients Univariable Analysis Multivariable Analysis Variables Hazard Ratio(95%CI) P Value Hazard Ratio(95%CI) P Value Age 0.981 (0.954 to 1.008) 0.162 Sex 0.363 male reference reference female 1.244 (0.778 to 1.989) 0.362 Smoking 0.064 no reference reference yes 0.637 (0.392 to 1.035) 0.069 Surgery way 0.004 thoracotomy reference reference reference reference thoracoscopy 0.499 (0.312 to 0.801) 0.004 0.845 (0.495 to 1.444) 0.538 Pathological stage 0.929 IIB reference reference IIIA 1.074 (0.649 to 1.779) 0.781 IIIB 0.945 (0.428 to 2.088) 0.889 T < 0.001 1 reference reference reference reference 2 4.263 (1.520 to 11.951) 0.006 3.195 (1.104 to 9.249) 0.032 3 2.533 (0.879 to 7.302) 0.085 2.824 (0.931 to 8.562) 0.067 4 0.917 (0.229 to 3.669) 0.903 0.815 (0.201 to 3.304) 0.775 N 0.185 0 reference reference 1 1.793 (0.905 to 3.550) 0.094 2 1.807 (1.012 to 3.227) 0.045 3 1.523 (0.352 to 6.592) 0.574 Histological type 0.002 squamous carcinoma reference reference reference reference adenocarcinoma 1.563 (0.745 to 3.278) 0.238 1.268 (0.576 to 2.792) 0.555 other 8.753 (3.016 to 25.402) < 0.001 10.661 (3.498 to 32.489) < 0.001 Tumor location 0.767 superior lobe of right lung reference reference middle lobe of right lung 1.275 (0.512 to 3.175) 0.602 inferior lobe of right lung 0.884 (0.458 to 1.706) 0.713 superior lobe of left lung 0.772 (0.412 to 1.447) 0.42 inferior lobe of left lung 0.710 (0.323 to 1.558) 0.393 Maximum tumor diameter 1.000 (0.885 to 1.131) 0.995 GGO 0.01 no reference reference reference reference yes 0.327 (0.119 to 0.898) 0.03 0.531 (0.188 to 1.501) 0.233 Air bronchogram 0.146 no reference reference yes 0.670 (0.384 to 1.171) 0.16 Lobulated 0.412 no reference reference yes 0.748 (0.383 to 1.462) 0.396 Spiculation < 0.001 no reference reference reference reference yes 2.936 (1.576 to 5.468) 0.001 2.464 (1.287 to 4.717) 0.007 Pleural indentation 0.075 no reference reference yes 1.613 (0.934 to 2.786) 0.086 Cavity 0.595 no reference reference yes 1.215 (0.603 to 2.447) 0.586 Vacuole 0.498 no reference reference yes 1.212 (0.702 to 2.093) 0.491 Emphysema 0.053 no reference reference yes 2.059 (1.053 to 4.025) 0.035 Pleural effusion 0.399 no reference reference yes 1.714 (0.539 to 5.452) 0.361 Other metastatic sites 0.001 no reference reference reference reference yes 2.717 (1.615 to 4.570) < 0.001 2.061 (1.186 to 3.581) 0.01 CEA 0.003 normal reference reference reference reference rise 2.186 (1.278 to 3.737) 0.004 1.650 (0.931 to 2.922) 0.086 GGO, ground-glass opacity; CEA, carcinoembryonic antigen; CI, confidence interval Model performance comparison The combined model exhibited significantly greater efficacy than both the radiomics and clinical models across the different machine learning algorithms. With respect to the training cohort, the QDA model demonstrated superior efficacy, with an AUC of 0.979 (95% CI, 0.964–0.995), accompanied by sensitivity and specificity values of 0.943 and 0.941, respectively. In the validation cohort, the AUC was 0.847 (95% CI, 0.743–0.952), with sensitivity and specificity values of 0.741 and 0.874, respectively (Table 3 ). The Supplemental Fig.S3 displays the p values of the ROC Delong test for the four machine learning algorithm models and the Cox proportional hazards risk model within the combined model. Notably, the AUC of the QDA model surpassed that of the other algorithms solely in the training cohort, and exhibited statistically significant differences (DeLong test, p = 0.03; p = 0.02; p = 0.02). However, there were no significant differences in the validation cohort (DeLong test, p > 0.05). With respect to the training cohort, all the machine learning models outperformed the Cox proportional hazards model. The calibration curves of the combined model showed strong agreement between the predicted and actual probabilities of BM in both the training and validation cohorts (Supplemental Fig.S4). The Hosmer–Lemeshow test confirmed that there was no difference between the predicted and actual probabilities of BM in either cohort, indicating excellent model fit. The KM survival curve (Supplemental Fig.S5) further verified the efficacy of the combined model. Figure 3 illustrates the DCA of the five models, which indicated that within a specific threshold probability range, the combined model offered greater net benefits than both the radiomics and clinical models for predicting BM. Model explanation The SHAP values can be used in SVM, GLM, LDA and QDA machine learning algorithms to rank the importance of features. From the perspective of global interpretation, the top features after the analysis of important radiomics features using these four machine learning algorithms are shown in Fig. 4 . For BM and without BM in the lung classification task, a positive SHAP value (yellow) indicates a BM marker, and a negative SHAP value (purple) indicates no BM marker. In addition, a nomogram was built based on the T stage, histological grade of the tumor, spicule, other metastatic status and rad-score, as shown in Supplemental Fig.S6a. Finally, we constructed a dynamic nomogram web calculator for clinicians for convenient use online at https://dynamic-nomogram-predict.shinyapps.io/DynNomapp/ . Due to the number of processes that shiny.app places on the instance, the instance cannot support concurrent running; thus, after accessing the instance, it will take at least five minutes to enter it again. Discussion In this study, machine learning technology was used for the first time in combination with the SHAP method, radiomics features and clinicopathological features to construct a risk prediction model for postoperative BM in patients with NSCLC. Finally, the model constructed with intra and peritumoral radiomics features and clinicopathological features exhibited good performance. The model was interpreted through SHAP values to determine which features were most closely associated with improved stability and to discover the most important features for predicting BM. Therefore, this feature of tumor images can reliably predict BM to help patients and clinicians make clinical decisions in an understandable way. Radiomics provides more detailed features than human eyes. our results show that CT radiomics features from the peritumor area can predict the risk of postoperative BM in NSCLC and that combined with intratumoral radiomics features, the efficacy of the model is greatly improved, with AUC ranging from 0.928 to 0.888 in the training cohort and from 0.894 to 0.838 in the validation cohort. It indicates that the prognosis of lung cancer is not only reflected in the lesion site, but also in the tissues around the tumor (Gao et al. 2023 ). The microenvironment surrounding the tumor is associated with aggressiveness (Braman et al. 2017 ; Chen et al. 2022 ), and capillaries and various cells around the tumor boundaries may be more active than those inside the tumor. The peritumoral characteristics were substantially related to tumor spread through air spaces status(Liao et al. 2022 ). Algohari et al.(2020) studied the density of peritumoral mesenchymal macrophages, epithelial cells and lymphocytes and found that it was associated with the risk of prostate cancer metastasis This finding suggested that the tissue surrounding the tumor is helpful in predicting patient prognosis. In addition, the model combining intratumoral and peritumoral radiomics features was more effective in predicting BM than the clinical model, which is consistent with previous findings. Zheng et al.(2023) found that the PET/CT radiomics model was more effective in predicting BM than the clinical model, with an AUC of 0.911 in the training cohort and 0.833 in the internal validation cohort. The radiomics model constructed by Ding et al.(Ding et al. 2022 ) and the combined model combining radiomics and clinical features were superior to the clinical model. The AUC of the radiomics model was 0.870 in the training cohort, 0.824 in the validation cohort, and the AUC of the combined model was 0.912 and 0.859 in the training cohort and validation cohort, respectively. The AUC of the clinical model was 0.712 in the training set and 0.692 in the validation set. Jian et al.(Jiang et al. 2022 ) established the radiomics model based on 8 selected features with a C-index of 0.733 (95%CI, 0.637–0.828) in the training cohort and 0.693 (95%CI, 0.569–0.818) in the validation cohort, which was also higher than the efficacy of the clinical model. The combined model of radiomics combined with clinical features constructed by Sun et al.(F. Sun et al. 2021 ) is superior to the clinical model, which is consistent with our research results. Radiomics specifically provides more complementary value to clinical information and improves the efficiency of the predictive model. The results of clinical risk factor analysis showed that the T stage of the tumor was an independent risk factor for BM, and it could be seen from the column diagram and SHAP diagram that the T4 stage was associated with a lower risk than was the other stages. This finding was different from the results of other studies that showed that the larger the tumor and the higher the T stage were, the greater the probability of BM(Zhang et al. 2023 ). The reason may be that most of the T4 stage patients included in this study had N stage N0 - N1, which made the overall stage relatively low. However, other studies have shown that T stage is not an independent risk factor for brain metastasis(F. Sun et al. 2021 ). In this study, there was no significant correlation between total stage and BM incidence. It should be carefully considered that this study included only resectable stage IIB - IIIB patients. In addition, histopathological type is also an independent risk factor for BM. Patients with nonsquamous cell carcinoma and adenocarcinoma are more likely to develop BM, and patients with adenocarcinoma are more likely to develop BM than are those with squamous cell carcinoma, which is consistent with the findings of other studies (F. Sun et al. 2021 ; S. Sun et al. 2021 ). Spicules are caused by lung cancer cells infiltrating adjacent normal lung tissue; pulmonary slippage is an important manifestation of borderline aggressiveness of malignant tumors and was considered an independent risk factor for BM in this study. Li et al.(2023) suggested that spicules have certain value in promoting angiogenesis in lung cancer patients, which may be related to distant metastasis. Previous predictive models for machine learning were difficult to interpret, therefore, we also used a new tool, SHAP, recently developed to interpret the results of “black box” machine learning models. In interpretability studies, the SHAP value for each feature is calculated. This approach provides great help for clinicians in understanding the model, and the model is more practical and generalizable. Deng et al.(2023) used a radiomics model combined with contrast-enhanced T1 MR, Xogboost, and SHAP algorithms and showed promise in accurately and interpretively identifying brain lesions in patients with NSCLC. In addition, the dynamic nomogram we used not only provides additional convenience for clinicians but also further advances scientific research into clinical practice. Limitations of this study. First, this was a single-center retrospective study, multi-center data will be needed in the future to prove the validity and accuracy of the model. Secondly, the data of some samples were incomplete, and the combined prediction of multiple omics, such as pathomics, genomics, etc., was not carried out. Third, the mining of CT images needs to be further improved, and the sample size is expected to be expanded for deep learning research in the future(Gu et al., 2022 ). Conclusions In conclusion, we developed a prediction model for the probability of BM in IIB - IIIB NSCLC by merging radiomics characteristics with clinicopathological data. Radiomics features may predict postoperative BM in IIB-IIIB NSCLC patients. The dynamic nomogram can help clinicians screen out high-risk groups of BM conveniently, with a view to providing support for treatment decisions. Declarations Acknowledgements We thank all the patients and the participating study teams for making this study possible. Statements and Declarations Funding This study is supported by the National Natural Science Foundation of China (82060313, 82160340), the Outstanding Youth Science Foundation of Yunnan Basic Research Project (202201AW070002), the Young and Middle-aged Academic and Technical leaders Reserve Talents Project of Yunnan (202305AC160020). Competing Interests The authors declare no potential conflicts of interest. Author Contributions CDL, YCH, and JY participated in the design of this study. LY, ZQYO, QQL performed data collection, region of interest segmentation and statistical analysis. ZQYO, QQL built predictive models and trains them. LY, CDL drafted the manuscript. All authors read and approved the final manuscript. Data Availability The datasets generated during during the current study are available from the corresponding author on reasonable request. Ethics approval This retrospective analysis was approved by the ethical review board of our hospital (No. KYLX-2023005). Consent to participate Informed consent was obtained from all individual participants included in the study. References Abbas YM, Khan MI (2023) Robust Machine Learning Framework for Modeling the Compressive Strength of SFRC: Database Compilation, Predictive Analysis, and Empirical Verification. Materials 16(22). https://doi.org/10.3390/ma16227178 Achrol AS, Rennert RC, Anders C et al (2019) Brain metastases. Nature Reviews Disease Primers 5(1). https://doi.org/10.1038/s41572-018-0055-y Algohary A, Shiradkar R, Pahwa S et al (2020) Combination of Peri-Tumoral and Intra-Tumoral Radiomic Features on Bi-Parametric MRI Accurately Stratifies Prostate Cancer Risk: A Multi-Site Study. Cancers 12(8). https://doi.org/10.3390/cancers12082200 He J, Wang X, Xiao R et al (2018) Risk factors for brain metastases in patients with non-small-cell lung cancer. Cancer Med 7(12), 6357–6364. https://doi.org/10.1002/cam4.1865 Angelini M, Blasilli G, Lenti S et al (2023) A Visual Analytics Conceptual Framework for Explorable and Steerable Partial Dependence Analysis. IEEE Transactions on Visualization and Computer Graphics 1–16. https://doi.org/10.1109/tvcg.2023.3263739 Braman NM, Etesami M, Prasanna P et al (2017) Intratumoral and peritumoral radiomics for the pretreatment prediction of pathological complete response to neoadjuvant chemotherapy based on breast DCE-MRI. Breast Cancer Research 19(1). https://doi.org/10.1186/s13058-017-0846-1 Chalubinska-Fendler J, Kepka L (2021) Prophylactic cranial irradiation in non-small cell lung cancer: evidence and future development. Journal of Thoracic Disease 13(5):3279–3288. https://doi.org/10.21037/jtd.2019.11.36 Chen Q, Shao J, Xue T et al (2022) Intratumoral and peritumoral radiomics nomograms for the preoperative prediction of lymphovascular invasion and overall survival in non-small cell lung cancer. European Radiology 33(2):947–958. https://doi.org/10.1007/s00330-022-09109-3 Deng F, Liu Z, Fang W et al (2023) MRI radiomics for brain metastasis sub-pathology classification from non-small cell lung cancer: a machine learning, multicenter study. Physical and Engineering Sciences in Medicine 46(3):1309–1320. https://doi.org/10.1007/s13246-023-01300-0 Dercle L, Fronheiser M, Lu L et al (2020) Identification of Non–Small Cell Lung Cancer Sensitive to Systemic Cancer Therapies Using Radiomics. Clinical Cancer Research 26(9):2151–2162. https://doi.org/10.1158/1078-0432.Ccr-19-2942 Ding Z, Wang Y, Xia C et al (2022) Thoracic CT radiomics analysis for predicting synchronous brain metastasis in patients with lung cancer. Diagnostic and Interventional Radiology 28(1):39–49. https://doi.org/10.5152/dir.2021.21677 Gao D, Fang L, Liu C et al (2023) Microenvironmental regulation in tumor progression: Interactions between cancer-associated fibroblasts and immune cells. Biomedicine & Pharmacotherapy 167. https://doi.org/10.1016/j.biopha.2023.115622 Gebreyesus Y, Dalton D, Nixon S et al (2023) Machine Learning for Data Center Optimizations: Feature Selection Using Shapley Additive exPlanation (SHAP). Future Internet 15(3). https://doi.org/10.3390/fi15030088 Gore EM, Bae K, Wong SJ et al (2011) Phase III Comparison of Prophylactic Cranial Irradiation Versus Observation in Patients With Locally Advanced Non–Small-Cell Lung Cancer: Primary Analysis of Radiation Therapy Oncology Group Study RTOG 0214. Journal of Clinical Oncology 29(3): 272–278. https://doi.org/10.1200/jco.2010.29.1609 Gu J, Tong T, Xu D et al (2022) Deep learning radiomics of ultrasonography for comprehensively predicting tumor and axillary lymph node status after neoadjuvant chemotherapy in breast cancer patients: A multicenter study. Cancer 129(3):356–366. https://doi.org/10.1002/cncr.34540 Hou Q, Sun B, Yao N et al (2022) Construction of Brain Metastasis Prediction Model and Optimization of Prophylactic Cranial Irradiation Selection for Limited-Stage Small-Cell Lung Cancer. Cancers 14(19). https://doi.org/10.3390/cancers14194906 Ji Z, Bi N, Wang J et al (2014) Risk factors for brain metastases in locally advanced non-small cell lung cancer with definitive chest radiation. Int J Radiat Oncol Biol Phys 89(2):330–337. https://doi.org/10.1016/j.ijrobp.2014.02.025 Jiang Y, Wang Y, Fu S et al (2022) A CT-based radiomics model to predict subsequent brain metastasis in patients with ALK‐rearranged non–small cell lung cancer undergoing crizotinib treatment. Thoracic Cancer 13(11):1558–1569. https://doi.org/10.1111/1759-7714.14386 Lambin P, Rios-Velazquez E, Leijenaar R et al (2012). Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer 48(4):441–446. https://doi.org/10.1016/j.ejca.2011.11.036 Dou TH, Coroller TP, van Griethuysen JJM (2018). Peritumoral radiomics features predict distant metastasis in locally advanced NSCLC. Plos One 13(11). https://doi.org/10.1371/journal.pone.0206108 Li S, Yang Z, Li Y et al (2023). Preoperative prediction of vasculogenic mimicry in lung adenocarcinoma using a CT radiomics model. Clinical Radiology. https://doi.org/10.1016/j.crad.2023.09.027 Liao G, Huang L, Wu S et al (2022). Preoperative CT-based peritumoral and tumoral radiomic features prediction for tumor spread through air spaces in clinical stage I lung adenocarcinoma. Lung Cancer 163:87–95. https://doi.org/10.1016/j.lungcan.2021.11.017 Lubner MG, Smith AD, Sandrasegaran K (2017) CT Texture Analysis: Definitions, Applications, Biologic Correlates, and Challenges. Radiographics, 37(5):1483–1503. https://doi.org/10.1148/rg.2017170056 Mitchell R, Frank E, Holmes G (2022) GPUTreeShap: massively parallel exact calculation of SHAP scores for tree ensembles. PeerJ Computer Science 8. https://doi.org/10.7717/peerj-cs.880 Ouyang W, Yu J, Zhou Y et al (2020) Risk factors of metachronous brain metastasis in patients with EGFR-mutated advanced non-small cell lung cancer. BMC Cancer 20(1):699. https://doi.org/10.1186/s12885-020-07202-8 Smith DR, Bian Y, Wu CC et al (2019) Natural history, clinical course and predictors of interval time from initial diagnosis to development of subsequent NSCLC brain metastases. J Neurooncol 143(1):145–155. https://doi.org/10.1007/s11060-019-03149-4 Sun F, Chen Y, Chen X et al (2021) CT-based radiomics for predicting brain metastases as the first failure in patients with curatively resected locally advanced non-small cell lung cancer. Eur J Radiol 134:109411. https://doi.org/10.1016/j.ejrad.2020.109411 Sun S, Men Y, Kang J et al (2021) A Nomogram for Predicting Brain Metastasis in IIIA-N2 Non-Small Cell Lung Cancer After Complete Resection: A Competing Risk Analysis. Frontiers in Oncology 11. https://doi.org/10.3389/fonc.2021.781340 Witlox WJA, Ramaekers BLT, Zindler JD et al (2018) The Prevention of Brain Metastases in Non-Small Cell Lung Cancer by Prophylactic Cranial Irradiation. Front Oncol 8:241. https://doi.org/10.3389/fonc.2018.00241 Zhang X, Gao H, Dang S et al (2023). Extracranial metastasis sites correlate to the incidence risk of brain metastasis in stage IV non-small cell lung cancer: a population-based study. Journal of Cancer Research and Clinical Oncology 149(9):6293–6301. https://doi.org/10.1007/s00432-022-04548-3 Zheng Z, Wang J, Tan W et al (2023) 18F-FDG PET/CT radiomics predicts brain metastasis in I-IIIA resected Non-Small cell lung cancer. European Journal of Radiology 165. https://doi.org/10.1016/j.ejrad.2023.110933 Zhou Y, Wang B, Qu J et al (2020) Survival outcomes and symptomatic central nervous system (CNS) metastasis in EGFR-mutant advanced non-small cell lung cancer without baseline CNS metastasis: Osimertinib vs. first-generation EGFR tyrosine kinase inhibitors. Lung Cancer 150:178–185. https://doi.org/10.1016/j.lungcan.2020.10.018 Additional Declarations No competing interests reported. Supplementary Files supplementarymaterials.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3972347","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":274119338,"identity":"ee898e31-9ce5-41ba-84e7-39948e343536","order_by":0,"name":"Lu yang","email":"","orcid":"","institution":"Yunnan Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Lu","middleName":"","lastName":"yang","suffix":""},{"id":274119339,"identity":"af00a269-e958-434d-8529-e3e9595d7fb3","order_by":1,"name":"Chengde Liao","email":"","orcid":"","institution":"Yan'an Hospital Affiliated To Kunming Medical University","correspondingAuthor":false,"prefix":"","firstName":"Chengde","middleName":"","lastName":"Liao","suffix":""},{"id":274119340,"identity":"42fdca4c-04c9-42b7-bcf7-b988b4e02268","order_by":2,"name":"Zhiqiang Ouyang","email":"","orcid":"","institution":"Yan'an Hospital Affiliated To Kunming Medical University","correspondingAuthor":false,"prefix":"","firstName":"Zhiqiang","middleName":"","lastName":"Ouyang","suffix":""},{"id":274119341,"identity":"d5119d61-358b-4a53-92ed-eeeadab2d911","order_by":3,"name":"Qinqing Li","email":"","orcid":"","institution":"Yunnan Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Qinqing","middleName":"","lastName":"Li","suffix":""},{"id":274119342,"identity":"52639997-92a5-49e1-9205-22d06017e58f","order_by":4,"name":"Yunchao Huang","email":"","orcid":"","institution":"Yunnan Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yunchao","middleName":"","lastName":"Huang","suffix":""},{"id":274119343,"identity":"b53534d2-222a-41fd-aa06-8e8c6fd121b8","order_by":5,"name":"Jun yang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAqElEQVRIiWNgGAWjYBACNgYGAxAtR7oWY5IsAmtJbCBaPZ/Y4W0SP3fUpvcdT2D88DGHGIdJp5VJ9p45njvzzANmyZnbiNKSYybB23Ysd8ONBDZmXmK1SP5tO5ZuQJIWad62mgRStKQVW8u2HTCceeZhM3F+kZ+dvPHm27Y6eb7jyQc/fCRGCxCwSDAwHGZgOEBC1DB/YGCoA2pJIFrHKBgFo2AUjDAAAFABNld46zLMAAAAAElFTkSuQmCC","orcid":"","institution":"Yunnan Cancer Hospital","correspondingAuthor":true,"prefix":"","firstName":"Jun","middleName":"","lastName":"yang","suffix":""}],"badges":[],"createdAt":"2024-02-20 09:31:51","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3972347/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3972347/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":51511869,"identity":"feee5889-4274-42b0-8105-31c2335c76ec","added_by":"auto","created_at":"2024-02-22 21:10:55","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":220982,"visible":true,"origin":"","legend":"\u003cp\u003eStudy flow chart. ROI, region of interest; SVM, support vector machine; GLM, generalize linear model; LDA, linear discriminant analysis; QDA, quadratic discriminant analysis; BM, brain metastasis.\u003c/p\u003e","description":"","filename":"Picture1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3972347/v1/32a4b9992410889b681f20a2.jpg"},{"id":51511871,"identity":"a8f7916e-9194-4356-875d-554968581f44","added_by":"auto","created_at":"2024-02-22 21:10:55","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":345965,"visible":true,"origin":"","legend":"\u003cp\u003eThe receiver operating characteristic curves of radiomics model 1, radiomics model 2, radiomics model 3, clinical model and combined model in training cohort (a, c, e, g, i) and validation cohort (b, d, f, h, j)\u003c/p\u003e","description":"","filename":"Picture2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3972347/v1/94c88a31601c8f2baff53314.jpg"},{"id":51511870,"identity":"457c8968-752f-4104-b271-549b118465ac","added_by":"auto","created_at":"2024-02-22 21:10:55","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":139085,"visible":true,"origin":"","legend":"\u003cp\u003eDecision curve analyses of radiomics model, clinical model and combined model in training cohort (a), validation cohort (b).\u003c/p\u003e","description":"","filename":"Picture3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3972347/v1/06c8f927fee693ee965dabc0.jpg"},{"id":51511872,"identity":"417043f9-c4e8-49a4-80b9-1e5a2aa4952c","added_by":"auto","created_at":"2024-02-22 21:10:55","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":308161,"visible":true,"origin":"","legend":"\u003cp\u003eThe relationship between the SHAP value and the importance of the feature in the combined model (a) SVM model, (b) GLM model, (c) LDA model, (d) QDA model. SVM, support vector machine; GLM, generalize linear model; LDA, linear discriminant analysis; QDA, quadratic discriminant analysis\u003c/p\u003e","description":"","filename":"Picture4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-3972347/v1/f529b47158cbd716cb50de00.jpg"},{"id":64199226,"identity":"4256598e-1d66-4d51-8bbc-7c129ef10cee","added_by":"auto","created_at":"2024-09-10 01:01:47","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2203195,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3972347/v1/11d42385-a185-4711-92ae-b46523697919.pdf"},{"id":51511873,"identity":"02c15896-c613-4e17-8b27-2dde37c61ee6","added_by":"auto","created_at":"2024-02-22 21:10:55","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":2225895,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-3972347/v1/9d5c4b63931d93d3a4ca55a8.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Risk prediction model for postoperative brain metastasis in IIB-IIIB non-small cell lung cancer: based on radiomics and clinicopathology","fulltext":[{"header":"Introduction","content":"\u003cp\u003eBrain metastases (BM) are the predominant distant metastases in non-small cell lung cancer (NSCLC)(Achrol et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Once a patient develops BM, the median progression-free survival is typically only 1\u0026ndash;2 months, and the one-year survival rate can reach as low as 10% \u0026minus;\u0026thinsp;20%(Smith et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), which poses a significant threat to patient survival and adversely affects quality of life(Ouyang et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Witlox et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Early prevention and diagnosis are crucial for improving patient outcomes. Prophylactic whole-brain irradiation has been proven effective at reducing the incidence of BM (Gore et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Witlox et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), but it may also lead to certain adverse reactions (Chalubinska et al. 2021). The challenge for clinicians lies in accurately weighing the benefits and drawbacks of radiotherapy while identifying high-risk patients who would benefit from early intervention(Chalubinska et al. 2021; Hou et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Therefore, precise prediction and stratification of BM can assist clinicians in implementing targeted preventive measures.\u003c/p\u003e \u003cp\u003eRadiomics, as a noninvasive method, enables the capture of features in CT images that are imperceptible to human eyes and quantitatively evaluate key information about the entire tumor, which can correlate these features with the prognosis of patients with cancer(Lambin et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Lubner et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). At present, many radiomics-based models have emerged to predict the prognosis of patients, including BM(An et al. 2018; F. Sun et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). However, the predictive efficiency of the model needs to be further improved, so it is urgent to further explore the factors to improve the predictive efficiency of the model. At present, some researchers have proposed that tumors can invade the adjacent lung parenchyma when they spread through vascular, lymphatic, or airway infiltration(Dercle et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Conversely, lung tissue around the tumor may change with the migration of tumor cells. Therefore, the evaluation of the lung tissue surrounding the tumor may have some significance for the occurrence of BM. However, these microscopic features are difficult to recognize with the naked eye. Therefore, the use of radiomics to extract and quantify the characteristics of the tissue around the tumor can help us analyze the value of the peritumor tissue. At present, some researchers have found that combined intra- and peritumoral radiomics features have demonstrated the ability to predict lymphovascular infiltration and overall survival in patients with NSCLC(Chen et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The peritumour radiomics features may offer more valuable information than visible tumor features for risk stratification of distant metastases in locally advanced NSCLC patients(Lee et al. 2018). Therefore, developing a radiomic model that explores peritumour features holds significant importance for precision oncology.\u003c/p\u003e \u003cp\u003eRadiomics can use different machine learning algorithms to build models, but the main reason that affects the use of models is the complexity of these models themselves, and their operation process is difficult to explain, which brings great challenges to model users. Currently, several interpretability methods in machine learning can be used to understand and interpret the predictions made by models, such as partial dependence graphs (PDP)(Abbas et al. 2023) and SHapley Additive exPlanations (SHAP)(Mitchell et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). One disadvantage of PDP is that struggle to illustrate relationships between multiple features and predictions(Angelini et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). SHAP values, rooted in game theory\u0026rsquo;s Shapley value concept, fairly measure features\u0026rsquo; contributions to predictions in machine learning(Gebreyesus et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). For each prediction sample, the model produces a predicted value. In addition, each feature of this sample will have a corresponding SHAP value, which indicates how the feature contributes to the prediction of the specified data point. Some factors will have a positive impact on the forecast probability, while others will have a negative impact on it. Calculating SHAP values improves interpretability and helps clinicians understand the output predicted by the model.\u003c/p\u003e \u003cp\u003eOur study aims to develop and validate intratumoral and peritumoral radiomics models, clinical models, and combined models predicting postoperative BM risk in NSCLC patients. Additionally, to better understand how radiomics features distinguish BM, we calculated the SHAP values for each feature. Our ultimate goal is to provide clinicians with an accurate probability of BM risk so that they can identify high-risk patients for early intervention.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatient\u003c/h2\u003e \u003cp\u003eThe retrospective analysis included 4803 NSCLC patients who underwent lung cancer resection at Yunnan Cancer Hospital between October 2015 and December 2019. The inclusion criteria for patients were as follows: 1) had NSCLC whose postoperative pathological stage was IIB-IIIB, 2) underwent chest CT examination at our hospital before surgery, 3) underwent brain MRI examination before surgery to exclude BM, and 4) underwent regular CT and brain MRI examination after surgery. The exclusion criteria for patients were as follows: 1) had other malignant tumors, 2) received neoadjuvant therapy, 3) had atelectasis and difficultly determining tumor boundaries, and 4) had less than 3 years of follow-up except for patients with BM. Finally, 333 patients with stage IIB-IIIB NSCLC were included, 97 with BM and 236 without BM. The patients were randomly divided into training (n\u0026thinsp;=\u0026thinsp;234) and validation (n\u0026thinsp;=\u0026thinsp;99) cohorts at a 7:3 ratio which was shown in Supplemental Fig.\u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e. This study was approved by the Ethics Committee of Yunnan Cancer Hospital, and the need for informed consent was waived.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eFollow-up\u003c/h2\u003e \u003cp\u003ePatients underwent craniocerebral MRI every 3\u0026ndash;6 months during the initial two postoperative years, followed by annual or symptom-driven assessments, with a minimum 3-year follow-up. We chose 3 years as the cutoff point for BM follow-up because previous studies have shown that the cumulative incidence of BM in NSCLC patients without baseline BM reaches a stable level 3 years after diagnosis(Ji et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Zhou et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), indicating that most NSCLC BM develop within 3 years of follow-up. The follow-up ended in December 2022. The endpoints of the study were considered to be BM, which was considered to occur when 1) MRI manifestations of BM were observed by two radiologists, 2) the patient presented obvious clinical symptoms, and the imaging findings of BM were observed by the radiologist; 3) the pathological diagnosis after intracranial space occupation was confirmed as BM; and 4) when the lesion was suspected to be BM increased or decreased in volume after treatment during MRI follow-up, the patient was diagnosed with BM.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eImage acquisition and CT semantic feature extraction\u003c/h2\u003e \u003cp\u003eThe chest CT was examined using a Siemens 128 scanner with the following acquisition parameters: layer thickness, 1.0 mm; tube voltage, 120 KV; and tube current, 160 mA. All of the images were reconstructed using a sharp core (b60) and a standard lung window (window width: 1600 HU; window level: -700 HU). Two radiologists with 8 and 3 years of experience independently evaluated the CT images. The CT semantic features included the tumor location, maximum tumor diameter, presence or absence of ground glass density, air bronchial signs, lobulation, spiculation, pleural indentation, vacuoles, cavities, emphysema, pleural effusion, and other metastatic sites (see Supplemental S-1 for specific definitions).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eCT Image segmentation and feature extraction and selection\u003c/h2\u003e \u003cp\u003eThe procedure of this study is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Two radiologists with 8 and 3 years of experience manually segmented the region of interest (ROI) and extracted radiomics features in 3D-Slicer (version 5.1.0. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.slicer.org\u003c/span\u003e\u003cspan address=\"https://www.slicer.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The peritumoral region of interest (ROI) extended outward by 3 mm along the edge of the primary tumor, and the outward expansion distances, including the ribs, chest wall muscle and soft tissue, were manually removed. The main feature categories extracted included shape features, first-order features, gray-level co-occurrence matrix (GLCM), gray-level run-length matrix (GLRLM), gray-level size zone matrix (GLSZM), neighboring gray tone difference matrix (NGTDM) and wavelet-based features. The intraclass correlation coefficient (ICC) was calculated to evaluate the stability and repeatability of feature extraction, and features with an inter-observer ICC greater than 0.8 were included in the study.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe included radiomics features were standardized. First, the training cohort data were standardized, and the standardized parameters were applied to the validation cohort. The standardized value Z for each patient-specific characteristic value X was calculated by the formula Z\u0026thinsp;=\u0026thinsp;X- mean/standard deviation. The mean and standard deviation were generated from the data in the training cohort. To avoid multicollinearity, correlation analysis was performed on radiomics features, and features with an absolute Pearson correlation coefficient greater than 0.85 were retained. The Least absolute shrinkage and selection operator (LASSO) was used to select the retained features, and the 10-fold cross-validation method was used to verify and select the best λ value.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eClinicopathological features\u003c/h2\u003e \u003cp\u003eThe clinicopathological factors analyzed included sex, smoking status, age at operation, mode of operation, pathological stage, pathological type, carcinoembryonic antigen (CEA) level, postoperative treatment, and postoperative metastasis. In the training cohort, univariate Cox analysis was used to analyze the differences between various factors in the presence and absence of BM, and variables with \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were included in multivariate Cox regression analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eModel construction and performance evaluation\u003c/h2\u003e \u003cp\u003eThe radiomics, clinicopathological and CT semantic features were ultimately selected to construct 5 models. The models were as follows: radiomics model 1 (intratumoral radiomics features), radiomics model 2 (3 mm peritumoral radiomics features), radiomics model 3 (intratumoral combined with peritumoral radiomics features), clinical model (clinicopathological features and CT semantic features), and combined model (radiomics features, clinicopathological features and CT semantic features). A linear combination of the screened features and their weight coefficients was used to construct a Rad-score for each patient. In addition, four commonly used machine learning algorithms, namely, support vector machine (SVM), generalized linear model (GLM), linear discriminant analysis (LDA) and quadratic discriminant analysis (QDA), are used to filter the best classifier. A nomogram and SHAP value graph were constructed. The performance of the model was evaluated by the area under the curve (AUC), sensitivity, specificity, calibration curve, and decision curve analysis (DCA), and the model was validated using validation cohorts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eStatistical analysis was performed using R software (version 4.3.1; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.r-project.org\u003c/span\u003e\u003cspan address=\"https://www.r-project.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The continuous variables subject to a normal distribution are expressed as the mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation, while those not subject to a normal distribution are expressed as the median and interquartile range (IQR). Chi-square tests were used to compare differences between categorical variables. The independent sample t test was used for continuous variables with a normal distribution. The clinicopathological features were screened by univariate and multivariate Cox regression analyses. The area under the curve (AUC), sensitivity and specificity were calculated to quantify the performance of the model. The AUCs were compared using the Delong test. A calibration curve was drawn to evaluate the goodness of fit of the model, and the Hosmer\u0026ndash;Lemeshow test was used to evaluate the consistency of the predicted probability with the actual observed value. Finally, DCA was performed to evaluate the clinical utility of the prediction model by quantifying the net benefit when considering different threshold probability capabilities. The risk stratification ability of the nomograms was evaluated using Kaplan‒Meier (KM) survival curves and log-rank tests. \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered to indicate statistical significance.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eClinical characteristics of patients\u003c/h2\u003e \u003cp\u003eA total of 333 patients with NSCLC were enrolled in this study (October 2015 to December 2019). We randomly divided 333 patients into a training cohort (n\u0026thinsp;=\u0026thinsp;234) and a validation cohort (n\u0026thinsp;=\u0026thinsp;99). The clinicopathological features and CT semantic features of the patients are shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Among the 333 patients, 202 (60.7%) were males and 131 (39.3%) were females. There were 97 patients with BM and 236 patients without BM. Across the cohort, the median follow-up time was 3.32 years (2.24, 4.49, IQR), and the median time to onset of BM was 1 year (0.63, 1.87, IQR). Moreover, there was no significant difference in the distribution of each feature between the training cohort and the validation cohort.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDemographic and clinical pathological characteristics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraining cohort\u003c/p\u003e \u003cp\u003e(N\u0026thinsp;=\u0026thinsp;234)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eValidation cohort\u003c/p\u003e \u003cp\u003e(N\u0026thinsp;=\u0026thinsp;99)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003cp\u003e(N\u0026thinsp;=\u0026thinsp;333)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge (years)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e55.2(\u0026plusmn;\u0026thinsp;8.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e56.0(\u0026plusmn;\u0026thinsp;9.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e55.4(\u0026plusmn;\u0026thinsp;8.96)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.453\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.144\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003emale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e136 (58.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66 (66.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e202 (60.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003efemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e98 (41.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e33 (33.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e131 (39.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSmoking\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.805\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e124 (53.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e51 (51.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e175 (52.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e110 (47.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e48 (48.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e158 (47.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSurgery way\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.793\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ethoracotomy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e98 (41.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e43 (43.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e141 (42.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ethoracoscopy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e136 (58.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e56 (56.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e192 (57.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePathological stage\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.337\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIIB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e87 (37.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e37 (37.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e124 (37.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIIIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e118 (50.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e55 (55.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e173 (52.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIIIB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29 (12.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7 (7.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e36 (10.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.208\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31 (13.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e19 (19.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e50 (15.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e82 (35.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31 (31.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e113 (33.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e87 (37.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e29 (29.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e116 (34.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e34 (14.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e20 (20.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e54 (16.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.429\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e82 (35.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e34 (34.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e116 (34.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e43 (18.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25 (25.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e68 (20.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e103 (44.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e39 (39.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e142 (42.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6 (2.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (1.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7 (2.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHistological type\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.909\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003esquamous carcinoma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e40 (17.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15 (15.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e55 (16.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eadenocarcinoma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e187 (79.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e81 (81.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e268 (80.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eother\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 (3.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3 (3.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e10 (3.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTumor location\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.228\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003esuperior lobe of right lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e61 (26.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e34 (34.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e95 (28.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003emiddle lobe of right lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14 (6.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5 (5.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e19 (5.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einferior lobe of right lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e53 (22.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13 (13.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e66 (19.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003esuperior lobe of left lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e70 (29.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e34 (34.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e104 (31.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einferior lobe of left lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e36 (15.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13 (13.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e49 (14.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMaximum tumor diameter (cm)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.53(\u0026plusmn;\u0026thinsp;1.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.16(\u0026plusmn;\u0026thinsp;2.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.42(\u0026plusmn;\u0026thinsp;1.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGGO\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.894\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e202 (86.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e86 (86.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e288 (86.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e32 (13.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13 (13.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e45 (13.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAir bronchogram\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.82\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e166 (70.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e69 (69.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e235 (70.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e68 (29.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30 (30.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e98 (29.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLobulated\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.511\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e27 (11.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9 (9.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e36 (10.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e207 (88.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e90 (90.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e297 (89.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSpiculation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.053\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e82 (35.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24 (24.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e106 (31.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e152 (65.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e75 (75.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e227 (68.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePleural indention\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.697\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e76 (32.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30 (30.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e106 (31.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e158 (67.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e69 (69.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e227 (68.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCavity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.404\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e208 (88.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e91 (91.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e299 (89.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e26 (11.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8 (8.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e34 (10.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVacuole\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.811\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e184 (78.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e79 (79.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e263 (79.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e50 (21.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e20 (20.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e70 (21.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEmphysema\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.545\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e213 (91.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e88 (88.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e301 (90.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21 (9.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11 (11.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e32 (9.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePleural effusion\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.366\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e228 (97.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e98 (99.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e326 (97.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6 (2.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (1.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7 (2.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOther metastatic sites\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.353\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e199 (85.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e88 (88.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e287 (86.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e35 (15.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11 (11.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e46 (13.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCEA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.177\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003enormal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e94 (40.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32 (32.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e126 (37.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003erise\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e140 (59.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e67 (67.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e207 (62.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFellow time (days)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1196.5(803\u0026ndash;1596)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1232(909-1675.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1212(818\u0026ndash;1627)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.378\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003eGGO, ground-glass opacity; CEA, carcinoembryonic antigen\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eRadiomics feature selection and radiomics model efficiency\u003c/h2\u003e \u003cp\u003eA total of 1316 features were extracted from the intratumoral ROI, and 1302 features were extracted from the ROI 3 mm per peritumor. A total of 601 ineligible features (those with an ICC less than 0.8) were excluded to ensure the stability and repeatability of the radiomics features during segmentation. To address collinearity among radiomics features, Pearson correlation analysis was used to determine that 1100 radiomics features were significantly correlated (r\u0026thinsp;\u0026gt;\u0026thinsp;0.85), and one of the features was retained. LASSO was used to select the best features, and the optimal λ values selected for intratumoral, peritumoral, and intratumoral combined peritumoral radiomics features were 0.02458, 0.02273, and 0.04359, respectively. Finally, the optimal λ values produced 15, 10, and 11 features, respectively, which were used to construct the radiomics model (Supplemental Fig.S2). The Rad-score calculation formula is reported in Supplemental S-2.\u003c/p\u003e \u003cp\u003eThree radiomics models were constructed using four machine learning classifiers. In addition, a rad-score based on radiomics features was constructed and validated. With respect to radiomics model 1, the model efficiency of the QDA machine learning algorithm was greater than that of the other algorithms in the training cohort; the AUC was 0.833 (95% CI, 0.779\u0026ndash;0.886), and the AUC was 0.646 (95% CI, 0.538\u0026ndash;0.754) in the validation cohort (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). With respect to radiomics model 2, the SVM and GLM algorithms had higher model efficiency than the other algorithms in the training cohort. The efficiency of radiomics model 3 was greater than that of radiomics models 1 and 2. With respect to the training cohort, the model based on the QDA machine learning algorithm had greater efficacy, with an AUC of 0.928 (95% CI, 0.891\u0026ndash;0.964) in the training cohort, the AUC was 0.838 (95% CI, 0.731\u0026ndash;0.945) in the validation cohort (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eModel performance comparison\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e \u003cp\u003eTraining cohort\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c10\" namest=\"c7\"\u003e \u003cp\u003eValidation cohort\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003emodel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSensitivity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSpecificity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAUC (95%CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eSensitivity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eSpecificity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003eAUC (95%CI)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSVM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.782\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.586\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.866\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.779 (0.714 to 0.843)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.727\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.722\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.736 (0.622 to 0.84)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.838\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.857\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.829\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.903 (0.859 to 0.946)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.798\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.806\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.832 (0.742 to 0.92)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.915\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.800\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.963\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.903 (0.854 to 0.952)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.838\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.875\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.864 (0.785 to 0.94)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClinical model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.684\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.714\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.671\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.719 (0.646 to 0.792)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.616\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.852\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.528\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.705 (0.599 to 0.81)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCombined model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.923\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.829\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.963\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.958 (0.934 to 0.983)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.899\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.944\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.926 (0.870 to 0.98)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGLM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.600\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.854\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.788 (0.724 to 0.852)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.747\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.704\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.764\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.736 (0.624 to 0.84)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.829\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.857\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.817\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.903 (0.861 to 0.944)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.833 (0.740 to 0.92)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.880\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.814\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.909\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.916 (0.872 to 0.961)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.859\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.889\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.882 (0.808 to 0.95)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClinical model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.684\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.671\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.689\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.718 (0.647 to 0.790)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.616\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.852\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.528\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.706 (0.603 to 0.80)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCombined model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.919\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.857\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.945\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.965 (0.944 to 0.986)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.899\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.944\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.918 (0.857 to 0.97)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLDA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.739\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.671\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.768\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.777 (0.712 to 0.841)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.657\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.815\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.597\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.695 (0.583 to 0.80)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.786\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.871\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.897 (0.855 to 0.939)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.788\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.815\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.830 (0.737 to 0.92)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.863\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.800\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.890\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.905 (0.860 to 0.950)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.828\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.815\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.833\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.894 (0.827 to 0.96)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClinical model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.684\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.671\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.689\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.718 (0.646 to 0.789)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.616\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.852\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.528\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.706 (0.603 to 0.80)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCombined model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.886\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.915\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.956 (0.931 to 0.982)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.869\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.815\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.889\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.923 (0.868 to 0.97)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eQDA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.731\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.871\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.671\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.833 (0.779 to 0.886)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.576\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.926\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.444\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.646 (0.538 to 0.75)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.838\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.714\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.890\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.898 (0.856 to 0.940)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.798\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.593\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.875\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.734 (0.606 to 0.86)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.872\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.814\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.896\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.928 (0.891 to 0.964)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.848\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.875\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.838 (0.731 to 0.94)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClinical model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.722\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.657\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.763 (0.699 to 0.828)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.576\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.852\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.472\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.695 (0.588 to 0.80)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCombined model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.927\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.943\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.921\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.979 (0.964 to 0.995)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.838\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.875\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.847 (0.743 to 0.95)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadscore\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.667\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.786\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.616\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.761 (0.696 to 0.826)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.616\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.926\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.735 (0.633 to 0.83)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.829\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.757\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.860\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.889 (0.844 to 0.935)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.828\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.815\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.833\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.862 (0.779 to 0.94)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model 3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.846\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.786\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.872\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.888 (0.842 to 0.935)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.838\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.852\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.833\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.886 (0.813 to 0.95)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClinical COX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClinical model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.756\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.614\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.817\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.783 (0.719 to 0.847)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.657\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.593\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.681\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.657 (0.540 to 0.77)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadscore clinical COX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCombined model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.880\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.914\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.866\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.951 (0.926 to 0.977)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e0.879\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.852\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.889\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.905 (0.834 to 0.97)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"10\"\u003eAUC, area under the curve; CI, confidence interval; SVM, support vector machine; GLM, generalize linear model; LDA, linear discriminant analysis; QDA, quadratic discriminant analysis\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eSelection of clinical features and efficacy of clinical models\u003c/h2\u003e \u003cp\u003eUnivariate and multivariate Cox regression analyses were used to evaluate the significance of these features in predicting BM. The results showed that T stage, pathological type, spiculation, and the presence of other metastatic sites before BM were independent predictors of BM (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e2\u003c/span\u003e). After these features were incorporated into the clinical model, the Cox proportional hazard regression model showed greater efficacy in the training cohort than in the validation cohort, with an AUC of 0.783 (95% CI, 0.719\u0026ndash;0.847) and an AUC of 0.657 (95% CI, 0.540\u0026ndash;0.774) (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eUni- and multivariable cox regression analyses in training cohort comprising IIB-IIIB non\u0026ndash;small cell lung cancer patients\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUnivariable Analysis\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMultivariable Analysis\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHazard Ratio(95%CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eP Value\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eHazard Ratio(95%CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eP Value\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.981 (0.954 to 1.008)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.162\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003emale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003efemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.244 (0.778 to 1.989)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.362\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSmoking\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.064\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.637 (0.392 to 1.035)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.069\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSurgery way\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ethoracotomy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ethoracoscopy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.499 (0.312 to 0.801)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.845 (0.495 to 1.444)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.538\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePathological stage\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.929\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIIB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIIIA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.074 (0.649 to 1.779)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.781\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIIIB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.945 (0.428 to 2.088)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.889\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.263 (1.520 to 11.951)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.006\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.195 (1.104 to 9.249)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.533 (0.879 to 7.302)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.085\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.824 (0.931 to 8.562)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.067\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.917 (0.229 to 3.669)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.903\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.815 (0.201 to 3.304)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.775\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.185\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.793 (0.905 to 3.550)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.094\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.807 (1.012 to 3.227)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.045\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.523 (0.352 to 6.592)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.574\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHistological type\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003esquamous carcinoma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eadenocarcinoma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.563 (0.745 to 3.278)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.238\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.268 (0.576 to 2.792)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.555\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eother\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8.753 (3.016 to 25.402)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e10.661 (3.498 to 32.489)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTumor location\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.767\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003esuperior lobe of right lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003emiddle lobe of right lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.275 (0.512 to 3.175)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.602\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einferior lobe of right lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.884 (0.458 to 1.706)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.713\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003esuperior lobe of left lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.772 (0.412 to 1.447)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einferior lobe of left lung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.710 (0.323 to 1.558)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.393\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMaximum tumor diameter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.000 (0.885 to 1.131)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.995\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGGO\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.327 (0.119 to 0.898)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.531 (0.188 to 1.501)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.233\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAir bronchogram\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.146\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.670 (0.384 to 1.171)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLobulated\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.412\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.748 (0.383 to 1.462)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.396\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSpiculation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.936 (1.576 to 5.468)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.464 (1.287 to 4.717)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePleural indentation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.075\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.613 (0.934 to 2.786)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.086\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCavity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.595\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.215 (0.603 to 2.447)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.586\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVacuole\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.498\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.212 (0.702 to 2.093)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.491\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEmphysema\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.053\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.059 (1.053 to 4.025)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.035\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePleural effusion\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.399\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.714 (0.539 to 5.452)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.361\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOther metastatic sites\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eyes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.717 (1.615 to 4.570)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.061 (1.186 to 3.581)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCEA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003enormal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003ereference\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003erise\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.186 (1.278 to 3.737)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.650 (0.931 to 2.922)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.086\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003eGGO, ground-glass opacity; CEA, carcinoembryonic antigen; CI, confidence interval\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eModel performance comparison\u003c/h2\u003e \u003cp\u003eThe combined model exhibited significantly greater efficacy than both the radiomics and clinical models across the different machine learning algorithms. With respect to the training cohort, the QDA model demonstrated superior efficacy, with an AUC of 0.979 (95% CI, 0.964\u0026ndash;0.995), accompanied by sensitivity and specificity values of 0.943 and 0.941, respectively. In the validation cohort, the AUC was 0.847 (95% CI, 0.743\u0026ndash;0.952), with sensitivity and specificity values of 0.741 and 0.874, respectively (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The Supplemental Fig.S3 displays the \u003cem\u003ep\u003c/em\u003e values of the ROC Delong test for the four machine learning algorithm models and the Cox proportional hazards risk model within the combined model. Notably, the AUC of the QDA model surpassed that of the other algorithms solely in the training cohort, and exhibited statistically significant differences (DeLong test, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.03; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.02; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.02). However, there were no significant differences in the validation cohort (DeLong test, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). With respect to the training cohort, all the machine learning models outperformed the Cox proportional hazards model. The calibration curves of the combined model showed strong agreement between the predicted and actual probabilities of BM in both the training and validation cohorts (Supplemental Fig.S4). The Hosmer\u0026ndash;Lemeshow test confirmed that there was no difference between the predicted and actual probabilities of BM in either cohort, indicating excellent model fit. The KM survival curve (Supplemental Fig.S5) further verified the efficacy of the combined model. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e illustrates the DCA of the five models, which indicated that within a specific threshold probability range, the combined model offered greater net benefits than both the radiomics and clinical models for predicting BM.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eModel explanation\u003c/h2\u003e \u003cp\u003eThe SHAP values can be used in SVM, GLM, LDA and QDA machine learning algorithms to rank the importance of features. From the perspective of global interpretation, the top features after the analysis of important radiomics features using these four machine learning algorithms are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. For BM and without BM in the lung classification task, a positive SHAP value (yellow) indicates a BM marker, and a negative SHAP value (purple) indicates no BM marker. In addition, a nomogram was built based on the T stage, histological grade of the tumor, spicule, other metastatic status and rad-score, as shown in Supplemental Fig.S6a. Finally, we constructed a dynamic nomogram web calculator for clinicians for convenient use online at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://dynamic-nomogram-predict.shinyapps.io/DynNomapp/\u003c/span\u003e\u003cspan address=\"https://dynamic-nomogram-predict.shinyapps.io/DynNomapp/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Due to the number of processes that shiny.app places on the instance, the instance cannot support concurrent running; thus, after accessing the instance, it will take at least five minutes to enter it again.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, machine learning technology was used for the first time in combination with the SHAP method, radiomics features and clinicopathological features to construct a risk prediction model for postoperative BM in patients with NSCLC. Finally, the model constructed with intra and peritumoral radiomics features and clinicopathological features exhibited good performance. The model was interpreted through SHAP values to determine which features were most closely associated with improved stability and to discover the most important features for predicting BM. Therefore, this feature of tumor images can reliably predict BM to help patients and clinicians make clinical decisions in an understandable way.\u003c/p\u003e \u003cp\u003eRadiomics provides more detailed features than human eyes. our results show that CT radiomics features from the peritumor area can predict the risk of postoperative BM in NSCLC and that combined with intratumoral radiomics features, the efficacy of the model is greatly improved, with AUC ranging from 0.928 to 0.888 in the training cohort and from 0.894 to 0.838 in the validation cohort. It indicates that the prognosis of lung cancer is not only reflected in the lesion site, but also in the tissues around the tumor (Gao et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The microenvironment surrounding the tumor is associated with aggressiveness (Braman et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Chen et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and capillaries and various cells around the tumor boundaries may be more active than those inside the tumor. The peritumoral characteristics were substantially related to tumor spread through air spaces status(Liao et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Algohari et al.(2020) studied the density of peritumoral mesenchymal macrophages, epithelial cells and lymphocytes and found that it was associated with the risk of prostate cancer metastasis This finding suggested that the tissue surrounding the tumor is helpful in predicting patient prognosis.\u003c/p\u003e \u003cp\u003eIn addition, the model combining intratumoral and peritumoral radiomics features was more effective in predicting BM than the clinical model, which is consistent with previous findings. Zheng et al.(2023) found that the PET/CT radiomics model was more effective in predicting BM than the clinical model, with an AUC of 0.911 in the training cohort and 0.833 in the internal validation cohort. The radiomics model constructed by Ding et al.(Ding et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and the combined model combining radiomics and clinical features were superior to the clinical model. The AUC of the radiomics model was 0.870 in the training cohort, 0.824 in the validation cohort, and the AUC of the combined model was 0.912 and 0.859 in the training cohort and validation cohort, respectively. The AUC of the clinical model was 0.712 in the training set and 0.692 in the validation set. Jian et al.(Jiang et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) established the radiomics model based on 8 selected features with a C-index of 0.733 (95%CI, 0.637\u0026ndash;0.828) in the training cohort and 0.693 (95%CI, 0.569\u0026ndash;0.818) in the validation cohort, which was also higher than the efficacy of the clinical model. The combined model of radiomics combined with clinical features constructed by Sun et al.(F. Sun et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) is superior to the clinical model, which is consistent with our research results. Radiomics specifically provides more complementary value to clinical information and improves the efficiency of the predictive model.\u003c/p\u003e \u003cp\u003eThe results of clinical risk factor analysis showed that the T stage of the tumor was an independent risk factor for BM, and it could be seen from the column diagram and SHAP diagram that the T4 stage was associated with a lower risk than was the other stages. This finding was different from the results of other studies that showed that the larger the tumor and the higher the T stage were, the greater the probability of BM(Zhang et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The reason may be that most of the T4 stage patients included in this study had N stage N0 - N1, which made the overall stage relatively low. However, other studies have shown that T stage is not an independent risk factor for brain metastasis(F. Sun et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In this study, there was no significant correlation between total stage and BM incidence. It should be carefully considered that this study included only resectable stage IIB - IIIB patients. In addition, histopathological type is also an independent risk factor for BM. Patients with nonsquamous cell carcinoma and adenocarcinoma are more likely to develop BM, and patients with adenocarcinoma are more likely to develop BM than are those with squamous cell carcinoma, which is consistent with the findings of other studies (F. Sun et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; S. Sun et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Spicules are caused by lung cancer cells infiltrating adjacent normal lung tissue; pulmonary slippage is an important manifestation of borderline aggressiveness of malignant tumors and was considered an independent risk factor for BM in this study. Li et al.(2023) suggested that spicules have certain value in promoting angiogenesis in lung cancer patients, which may be related to distant metastasis.\u003c/p\u003e \u003cp\u003ePrevious predictive models for machine learning were difficult to interpret, therefore, we also used a new tool, SHAP, recently developed to interpret the results of \u0026ldquo;black box\u0026rdquo; machine learning models. In interpretability studies, the SHAP value for each feature is calculated. This approach provides great help for clinicians in understanding the model, and the model is more practical and generalizable. Deng et al.(2023) used a radiomics model combined with contrast-enhanced T1 MR, Xogboost, and SHAP algorithms and showed promise in accurately and interpretively identifying brain lesions in patients with NSCLC. In addition, the dynamic nomogram we used not only provides additional convenience for clinicians but also further advances scientific research into clinical practice.\u003c/p\u003e \u003cp\u003eLimitations of this study. First, this was a single-center retrospective study, multi-center data will be needed in the future to prove the validity and accuracy of the model. Secondly, the data of some samples were incomplete, and the combined prediction of multiple omics, such as pathomics, genomics, etc., was not carried out. Third, the mining of CT images needs to be further improved, and the sample size is expected to be expanded for deep learning research in the future(Gu et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn conclusion, we developed a prediction model for the probability of BM in IIB - IIIB NSCLC by merging radiomics characteristics with clinicopathological data. Radiomics features may predict postoperative BM in IIB-IIIB NSCLC patients. The dynamic nomogram can help clinicians screen out high-risk groups of BM conveniently, with a view to providing support for treatment decisions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e We thank all the patients and the participating study teams for making this study possible.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatements and Declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study is supported by the National Natural Science Foundation of China (82060313, 82160340), the Outstanding Youth Science Foundation of Yunnan Basic Research Project (202201AW070002), the Young and Middle-aged Academic and Technical leaders Reserve Talents Project of Yunnan (202305AC160020).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no potential conflicts of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCDL, YCH, and JY participated in the design of this study. LY, ZQYO, QQL performed data collection, region of interest segmentation and statistical analysis. \u0026nbsp; ZQYO, QQL built predictive models and trains them. LY, CDL drafted the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated during during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis retrospective analysis was approved by the ethical review board of our hospital (No. KYLX-2023005).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInformed consent was obtained from all individual participants included in the study.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbbas YM, Khan MI (2023) Robust Machine Learning Framework for Modeling the Compressive Strength of SFRC: Database Compilation, Predictive Analysis, and Empirical Verification. Materials 16(22). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ma16227178\u003c/span\u003e\u003cspan address=\"10.3390/ma16227178\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAchrol AS, Rennert RC, Anders C et al (2019) Brain metastases. Nature Reviews Disease Primers 5(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41572-018-0055-y\u003c/span\u003e\u003cspan address=\"10.1038/s41572-018-0055-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlgohary A, Shiradkar R, Pahwa S et al (2020) Combination of Peri-Tumoral and Intra-Tumoral Radiomic Features on Bi-Parametric MRI Accurately Stratifies Prostate Cancer Risk: A Multi-Site Study. Cancers 12(8). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/cancers12082200\u003c/span\u003e\u003cspan address=\"10.3390/cancers12082200\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe J, Wang X, Xiao R et al (2018) Risk factors for brain metastases in patients with non-small-cell lung cancer. Cancer Med 7(12), 6357\u0026ndash;6364. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/cam4.1865\u003c/span\u003e\u003cspan address=\"10.1002/cam4.1865\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAngelini M, Blasilli G, Lenti S et al (2023) A Visual Analytics Conceptual Framework for Explorable and Steerable Partial Dependence Analysis. IEEE Transactions on Visualization and Computer Graphics 1\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/tvcg.2023.3263739\u003c/span\u003e\u003cspan address=\"10.1109/tvcg.2023.3263739\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBraman NM, Etesami M, Prasanna P et al (2017) Intratumoral and peritumoral radiomics for the pretreatment prediction of pathological complete response to neoadjuvant chemotherapy based on breast DCE-MRI. Breast Cancer Research 19(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13058-017-0846-1\u003c/span\u003e\u003cspan address=\"10.1186/s13058-017-0846-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChalubinska-Fendler J, Kepka L (2021) Prophylactic cranial irradiation in non-small cell lung cancer: evidence and future development. Journal of Thoracic Disease 13(5):3279\u0026ndash;3288. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21037/jtd.2019.11.36\u003c/span\u003e\u003cspan address=\"10.21037/jtd.2019.11.36\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen Q, Shao J, Xue T et al (2022) Intratumoral and peritumoral radiomics nomograms for the preoperative prediction of lymphovascular invasion and overall survival in non-small cell lung cancer. European Radiology 33(2):947\u0026ndash;958. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00330-022-09109-3\u003c/span\u003e\u003cspan address=\"10.1007/s00330-022-09109-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeng F, Liu Z, Fang W et al (2023) MRI radiomics for brain metastasis sub-pathology classification from non-small cell lung cancer: a machine learning, multicenter study. Physical and Engineering Sciences in Medicine 46(3):1309\u0026ndash;1320. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s13246-023-01300-0\u003c/span\u003e\u003cspan address=\"10.1007/s13246-023-01300-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDercle L, Fronheiser M, Lu L et al (2020) Identification of Non\u0026ndash;Small Cell Lung Cancer Sensitive to Systemic Cancer Therapies Using Radiomics. Clinical Cancer Research 26(9):2151\u0026ndash;2162. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1158/1078-0432.Ccr-19-2942\u003c/span\u003e\u003cspan address=\"10.1158/1078-0432.Ccr-19-2942\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDing Z, Wang Y, Xia C et al (2022) Thoracic CT radiomics analysis for predicting synchronous brain metastasis in patients with lung cancer. Diagnostic and Interventional Radiology 28(1):39\u0026ndash;49. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5152/dir.2021.21677\u003c/span\u003e\u003cspan address=\"10.5152/dir.2021.21677\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGao D, Fang L, Liu C et al (2023) Microenvironmental regulation in tumor progression: Interactions between cancer-associated fibroblasts and immune cells. Biomedicine \u0026amp; Pharmacotherapy 167. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.biopha.2023.115622\u003c/span\u003e\u003cspan address=\"10.1016/j.biopha.2023.115622\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGebreyesus Y, Dalton D, Nixon S et al (2023) Machine Learning for Data Center Optimizations: Feature Selection Using Shapley Additive exPlanation (SHAP). Future Internet 15(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/fi15030088\u003c/span\u003e\u003cspan address=\"10.3390/fi15030088\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGore EM, Bae K, Wong SJ et al (2011) Phase III Comparison of Prophylactic Cranial Irradiation Versus Observation in Patients With Locally Advanced Non\u0026ndash;Small-Cell Lung Cancer: Primary Analysis of Radiation Therapy Oncology Group Study RTOG 0214. Journal of Clinical Oncology 29(3): 272\u0026ndash;278. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1200/jco.2010.29.1609\u003c/span\u003e\u003cspan address=\"10.1200/jco.2010.29.1609\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu J, Tong T, Xu D et al (2022) Deep learning radiomics of ultrasonography for comprehensively predicting tumor and axillary lymph node status after neoadjuvant chemotherapy in breast cancer patients: A multicenter study. Cancer 129(3):356\u0026ndash;366. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/cncr.34540\u003c/span\u003e\u003cspan address=\"10.1002/cncr.34540\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHou Q, Sun B, Yao N et al (2022) Construction of Brain Metastasis Prediction Model and Optimization of Prophylactic Cranial Irradiation Selection for Limited-Stage Small-Cell Lung Cancer. Cancers 14(19). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/cancers14194906\u003c/span\u003e\u003cspan address=\"10.3390/cancers14194906\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJi Z, Bi N, Wang J et al (2014) Risk factors for brain metastases in locally advanced non-small cell lung cancer with definitive chest radiation. Int J Radiat Oncol Biol Phys 89(2):330\u0026ndash;337. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ijrobp.2014.02.025\u003c/span\u003e\u003cspan address=\"10.1016/j.ijrobp.2014.02.025\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang Y, Wang Y, Fu S et al (2022) A CT-based radiomics model to predict subsequent brain metastasis in patients with ALK‐rearranged non\u0026ndash;small cell lung cancer undergoing crizotinib treatment. Thoracic Cancer 13(11):1558\u0026ndash;1569. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1759-7714.14386\u003c/span\u003e\u003cspan address=\"10.1111/1759-7714.14386\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLambin P, Rios-Velazquez E, Leijenaar R et al (2012). Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer 48(4):441\u0026ndash;446. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ejca.2011.11.036\u003c/span\u003e\u003cspan address=\"10.1016/j.ejca.2011.11.036\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDou TH, Coroller TP, van Griethuysen JJM (2018). Peritumoral radiomics features predict distant metastasis in locally advanced NSCLC. Plos One 13(11). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0206108\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0206108\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi S, Yang Z, Li Y et al (2023). Preoperative prediction of vasculogenic mimicry in lung adenocarcinoma using a CT radiomics model. Clinical Radiology. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.crad.2023.09.027\u003c/span\u003e\u003cspan address=\"10.1016/j.crad.2023.09.027\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiao G, Huang L, Wu S et al (2022). Preoperative CT-based peritumoral and tumoral radiomic features prediction for tumor spread through air spaces in clinical stage I lung adenocarcinoma. Lung Cancer 163:87\u0026ndash;95. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.lungcan.2021.11.017\u003c/span\u003e\u003cspan address=\"10.1016/j.lungcan.2021.11.017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLubner MG, Smith AD, Sandrasegaran K (2017) CT Texture Analysis: Definitions, Applications, Biologic Correlates, and Challenges. Radiographics, 37(5):1483\u0026ndash;1503. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1148/rg.2017170056\u003c/span\u003e\u003cspan address=\"10.1148/rg.2017170056\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMitchell R, Frank E, Holmes G (2022) GPUTreeShap: massively parallel exact calculation of SHAP scores for tree ensembles. PeerJ Computer Science 8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7717/peerj-cs.880\u003c/span\u003e\u003cspan address=\"10.7717/peerj-cs.880\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOuyang W, Yu J, Zhou Y et al (2020) Risk factors of metachronous brain metastasis in patients with EGFR-mutated advanced non-small cell lung cancer. BMC Cancer 20(1):699. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12885-020-07202-8\u003c/span\u003e\u003cspan address=\"10.1186/s12885-020-07202-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith DR, Bian Y, Wu CC et al (2019) Natural history, clinical course and predictors of interval time from initial diagnosis to development of subsequent NSCLC brain metastases. J Neurooncol 143(1):145\u0026ndash;155. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11060-019-03149-4\u003c/span\u003e\u003cspan address=\"10.1007/s11060-019-03149-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun F, Chen Y, Chen X et al (2021) CT-based radiomics for predicting brain metastases as the first failure in patients with curatively resected locally advanced non-small cell lung cancer. Eur J Radiol 134:109411. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ejrad.2020.109411\u003c/span\u003e\u003cspan address=\"10.1016/j.ejrad.2020.109411\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun S, Men Y, Kang J et al (2021) A Nomogram for Predicting Brain Metastasis in IIIA-N2 Non-Small Cell Lung Cancer After Complete Resection: A Competing Risk Analysis. Frontiers in Oncology 11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fonc.2021.781340\u003c/span\u003e\u003cspan address=\"10.3389/fonc.2021.781340\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWitlox WJA, Ramaekers BLT, Zindler JD et al (2018) The Prevention of Brain Metastases in Non-Small Cell Lung Cancer by Prophylactic Cranial Irradiation. Front Oncol 8:241. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fonc.2018.00241\u003c/span\u003e\u003cspan address=\"10.3389/fonc.2018.00241\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang X, Gao H, Dang S et al (2023). Extracranial metastasis sites correlate to the incidence risk of brain metastasis in stage IV non-small cell lung cancer: a population-based study. Journal of Cancer Research and Clinical Oncology 149(9):6293\u0026ndash;6301. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00432-022-04548-3\u003c/span\u003e\u003cspan address=\"10.1007/s00432-022-04548-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng Z, Wang J, Tan W et al (2023) 18F-FDG PET/CT radiomics predicts brain metastasis in I-IIIA resected Non-Small cell lung cancer. European Journal of Radiology 165. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ejrad.2023.110933\u003c/span\u003e\u003cspan address=\"10.1016/j.ejrad.2023.110933\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou Y, Wang B, Qu J et al (2020) Survival outcomes and symptomatic central nervous system (CNS) metastasis in EGFR-mutant advanced non-small cell lung cancer without baseline CNS metastasis: Osimertinib vs. first-generation EGFR tyrosine kinase inhibitors. Lung Cancer 150:178\u0026ndash;185. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.lungcan.2020.10.018\u003c/span\u003e\u003cspan address=\"10.1016/j.lungcan.2020.10.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Radiomics, Non-small cell lung cancer, Brain metastasis, Machine learning, Dynamic nomogram","lastPublishedDoi":"10.21203/rs.3.rs-3972347/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3972347/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003ePurpose\u003c/h2\u003e \u003cp\u003eTo develop and validate a model based on radiomics and clinicopathological features for predicting postoperative brain metastasis (BM) in stage IIB-IIIB non-small cell lung cancer (NSCLC) patients.\u003c/p\u003e\u003ch2\u003eMaterials and methods\u003c/h2\u003e \u003cp\u003eA total of 333 NSCLC patients operated from October 2015 and December 2019 with postoperative pathological stage IIB-IIIB were included, which were randomly divided into a training and validation cohort. The intratumoral and peritumoral radiomics features from preoperative CT image were extracted and selected using the least absolute shrinkage and selection operator (LASSO). The independent clinical predictors of BM were identified by univariate and multivariate Cox analysis. The radiomics model, clinical model and radiomics combined clinicopathological model were constructed with six different algorithms. Subsequently, we constructed a dynamic nomogram. The performance of the model was evaluated by the area under the curve (AUC), sensitivity, specificity, calibration curve and decision curve analysis (DCA).\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThe radiomics model combining intratumoral and peritumoral radiomics features exhibited great predictive performance for BM prediction, with an AUC of 0.888\u0026ndash;0.928 in the training cohort and 0.838\u0026ndash;0.894 in the validation cohort. The model including the intra- and peritumoral radiomics, T stage, histological type, spiculation and other metastatic sites yielded AUC of 0.947\u0026ndash;0.979 in the training cohort and 0.847\u0026ndash;0.926 in the validation cohort, with good calibration for all algorithms (p\u0026thinsp;\u0026gt;\u0026thinsp;0.05). DCA revealed that the combined model obtained a greater net benefit.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThe model that integrates radiomics features with clinicopathological features could aid in early-stage prediction of postoperative BM risk in stage IIB-IIIB NSCLC patients. Dynamic nomogram provides great convenience for clinicians to manage patients.\u003c/p\u003e","manuscriptTitle":"Risk prediction model for postoperative brain metastasis in IIB-IIIB non-small cell lung cancer: based on radiomics and clinicopathology","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-02-22 21:10:50","doi":"10.21203/rs.3.rs-3972347/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"5c5b928b-6188-41b1-9440-d1718d72c936","owner":[],"postedDate":"February 22nd, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-05-06T08:08:38+00:00","versionOfRecord":[],"versionCreatedAt":"2024-02-22 21:10:50","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3972347","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3972347","identity":"rs-3972347","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}