Dual-layer spectral detector CT quantitative parameters and radiomics for predicting spread through air spaces of lung adenocarcinoma: A dual-center study

Dual-layer spectral detector CT quantitative parameters and radiomics for predicting spread through air spaces of lung adenocarcinoma: A dual-center study | 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 Dual-layer spectral detector CT quantitative parameters and radiomics for predicting spread through air spaces of lung adenocarcinoma: A dual-center study Pei Huang, Ze Lin, Yingying Qiu, Dan Chen, Yiqing Tan, Li Fan, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7391071/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 22 Dec, 2025 Read the published version in BMC Cancer → Version 1 posted 9 You are reading this latest preprint version Abstract Objectives To evaluate the value of quantitative parameters and radiomic features based on dual-layer spectral detector CT (DLCT) in predicting spread through air spaces (STAS) of lung adenocarcinoma (LUAD). Methods This study analyzed 266 patients with pathologically confirmed LUAD from two medical centers. Patients from center 1 were divided into training (n = 131) and internal validation (n = 57) sets, while center 2 (n = 78) formed the external validation set. Clinical data, conventional imaging features, and DLCT quantitative parameters were analyzed to develop a clinical-radiological model. Radiomic features were extracted from venous-phase images, including conventional images, virtual monoenergetic images (VMI) at 40keV, 65keV, and 100keV, along with iodine density maps, effective atomic number (Zeff) maps, and electron density (ED) maps. The best-performing radiomics model was combined with clinical-radiological predictors to create a nomogram. Model performance was evaluated through ROC analysis, calibration curves, and decision curve analysis. Results Multivariate analysis revealed that tumor-lung interface and ED values were independent predictive factors in the clinical-radiological model. The optimal radiomics model was constructed based on VMI 40keV, demonstrating AUCs of 0.899, 0.835, and 0.828 in the training, internal validation, and external validation sets, respectively. The nomogram, which incorporated the VMI 40keV radiomics signature along with tumor-lung interface and ED values, outperformed the clinical-radiological model in the training set (AUC = 0.910 vs 0.870; P = 0.018) and the internal validation set (AUC = 0.868 vs 0.798; P = 0.046). While the improvement in the external validation set was not statistically significant (AUC = 0.848 vs 0.819; P = 0.184). Conclusion The nomogram, which integrates conventional imaging features, DLCT quantitative parameters and VMI 40keV radiomic features, serves as a valuable non-invasive tool for the preoperative assessment of STAS in LUAD. Dual-layer spectral computed tomography Spread through air spaces Lung adenocarcinoma Radiomics Nomogram Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction The 2015 World Health Organization classification introduced tumor spread through air spaces (STAS) as a novel lung cancer dissemination pattern, defined as micropapillary clusters, solid nests, or single tumor cells in air spaces beyond the tumor boundary[ 1 ]. This indicates that lung cancer can exhibit microscopic spread via air spaces. Current research on STAS predominantly focuses on lung adenocarcinoma (LUAD), the most common histologic subtype of lung cancer[ 2 , 3 ]. Several studies have suggested that certain STAS cases may be false positives caused by artifacts during surgery or pathological specimen preparation[ 4 , 5 ]. However, substantial evidence confirms STAS as an independent prognostic factor associated with reduced survival and increased recurrence[ 6 , 7 ]. For example, a previous study demonstrated that lobectomy yields superior survival outcomes compared with sublobar resection in stage I lung adenocarcinoma with STAS[ 8 ]. In contrast, sublobar resection remains appropriate for early-stage patients with non-invasive histologic features, as it preserves pulmonary parenchyma and improves postoperative lung function and quality of life[ 9 ]. Furthermore, accurate STAS identification is essential for determining optimal margins in patients undergoing stereotactic radiotherapy, which critically impacts local recurrence [ 10 ]. Postoperative histopathological diagnosis is the gold standard for confirming STAS. Currently, there is no reliable method for detecting STAS through preoperative biopsy or intraoperative frozen section[ 11 , 12 ], which may limit its substantial impact on treatment decisions. Therefore, the development of a non-invasive, simple, and practical method to accurately predict STAS preoperatively could provide crucial information to guide clinical decision-making. In clinical practice, owing to its superior resolution and cost-effectiveness, high-resolution CT has become the preferred preoperative imaging method for pulmonary lesions. Previous research has demonstrated a correlation between some conventional imaging characteristics of LUAD and the presence of STAS[ 2 , 13 ], but these features are largely dependent on the subjective experience of the interpreter, which may lead to potential bias. Recent advances in dual-layer spectral detector CT (DLCT) have transformed CT diagnosis from qualitative to quantitative assessment[ 14 , 15 ]. Beyond conventional images, DLCT produces energy-specific data such as virtual monoenergetic images (VMI), electron density (ED), and iodine density (ID) maps, enhancing diagnostic accuracy and clinical utility[ 15 ]. Many studies have shown that quantitative parameters from these images provide more accurate and reliable evidence for the differentiation of pulmonary lesions[ 16 – 18 ]. Additionally, radiomics enables the non-invasive assessment of tumor heterogeneity by rapidly extracting quantitative features from medical images[ 19 ]. These features have demonstrated substantial efficacy in differentiating benign from malignant pulmonary lesions, categorizing molecular subtypes, and evaluating prognostic[ 20 , 21 ]. However, the majority of radiomics research has focused on conventional CT imaging, with the potential usefulness of radiomic features from DLCT images remaining under-explored. To the best of our knowledge, few studies have investigated whether integrating these methods could provide additional benefits for evaluating pulmonary lesions. Therefore, this study aims to develop a nomogram that integrates clinical data, conventional imaging characteristics, DLCT quantitative parameters, and radiomic features for preoperative prediction of STAS in LUAD. Materials and methods Patients This retrospective study was conducted with the approval of the institutional ethics review board, which granted a waiver for written informed consent. Between September 2023 and July 2024, patients who underwent thoracic contrast-enhanced DLCT scans at Center 1 were included. After applying the specified inclusion and exclusion criteria, the eligible patients were randomly divided into training and internal validation sets at a 7:3 ratio. Similarly, between January 2024 and April 2024, patients from Center 2 who satisfied the same criteria were enrolled to establish the external validation set. Inclusion criteria: (1) postoperative pathology confirmed LUAD, and the status of STAS was determined; (2) initial diagnosis with no prior history of anti-tumor treatment; (3) DLCT scan performed within two weeks before surgery. Exclusion criteria: (1) postoperative pathological diagnosis of adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), or other special types such as mucinous, fetal, or intestinal-type adenocarcinoma; (2) images exhibiting severe artifacts, atelectasis, or obstructive pneumonia that obscures tumor margins; (3) incomplete clinical or pathological data; and (4) missing spectral-based images (SBI) data. The inclusion and exclusion flowchart for patient selection is shown in Figure 1 . DLCT scan protocol DLCT scans were performed using a 256-slice Hawk Spectral CT at Center 1 and a 64-slice IQon Spectral CT at Center 2 (both Philips Healthcare, Netherlands). Detailed information on the DLCT scanning protocol and contrast agent injection scheme is provided in Supplementary Table 1 . Automatic contrast agent tracking was employed with a threshold of 150 HU. The arterial phase scan was initiated with an 8-second delay, followed by the venous phase scan 35 seconds after the arterial phase. Following the scan, conventional 120 kVp images and SBI were reconstructed with a 1 mm slice thickness and 1 mm increment, including virtual monoenergetic images (VMI) 40keV, 65keV, and 100keV, iodine density (ID) maps, effective atomic number (Zeff) maps, and electron density (ED) maps. Clinical data collection and image analysis Clinical information, such as age, gender, and smoking history, was collected for all patients. Tumor conventional imaging features were independently reviewed by two radiologists (3 and 8 years of chest imaging diagnosis experience) while blinded to pathology. Disagreements were resolved through discussion to reach a consensus. Evaluated conventional radiological features included long-axis diameter (LD), consolidation/tumor ratio (CTR), shape, nodule type (pGGN, mixed solid, solid nodule), lobulation, spiculation, air bronchogram, tumor-lung interface, vacuole sign, and pleural indentation. DLCT quantitative parameter measurement On Philips post-processing workstation (IntelliSpace Portal, Philips Healthcare), DLCT quantitative parameters were independently measured by two radiologists. Final values were calculated as the average of both measurements. The largest axial plane of the lesion was identified on conventional 120 kVp lung window images (window width 1600 HU, window level -600 HU) in the venous-phase. The region of interest (ROI) for the tumor was then manually outlined along the lesion contour. Large blood vessels, calcification, necrosis, and cavities were carefully avoided when defining the ROI. To maintain consistency across all energy images, the workstation's "copy and paste" function was utilized. Measured quantitative parameters include spectral curve slope (λHU), normalized iodine density (NID), Zeff, and ED value. Additionally, to calculate NID, the ID of the aorta was also recorded at the same axial plane as the tumor. Supplemental Figure 1 presents a schematic for DLCT quantitative parameter measurement. λHU and NID were calculated as: λHU = (CT 40keV - CT 100keV )/100 - 40 NID = ID tumor /ID aorta Tumor delineation and radiomics feature extraction A radiologist manually delineated the three-dimensional ROI of the tumor on the conventional venous-phase images via ITK-SNAP (version 3.8.0). Radiomic features were extracted from seven type images: conventional, VMI at 40keV, 65keV, and 100keV, ID, Zeff, and ED. Since all DLCT energy images had the same dimensions, tumor ROI delineation was performed solely on the conventional image, which could be perfectly aligned with the reconstructed energy images. All images were standardized to a voxel size of 1 × 1 × 1 mm³, and a bin width of 25 was implemented to ensure uniform intensity scaling. The Pyradiomics package (version 3.6.0) was used to extract radiomic features, including first-order, shape, and texture features. Detailed information of these features is provided in Supplemental Appendix S1 . Additionally, another radiologist randomly selected 30 patients to delineate the ROI and assess inter-observer reproducibility by calculating the inter-class correlation coefficient (ICC). Feature selection and model construction Radiomic features with an ICC ≥ 0.75, reflecting good reproducibility, were retained for further analysis. These features were subsequently standardized to Z-scores to ensure comparability across different scales. A three-step approach was employed to refine feature selection within the training dataset. First, features with P -value < 0.05 were selected using the Mann-Whitney U test. Second, to address collinearity among features, Pearson's correlation analysis was performed. When the correlation coefficient between two features exceeded 0.9, one of the correlated features was excluded. Finally, the least absolute shrinkage and selection operator (LASSO) algorithm, with 5-fold cross-validation, was applied to identify the optimal feature set. Based on these selected features from each type image, radiomics models were constructed using a logistic regression (LR) machine learning classifier, resulting in seven distinct radiomic signatures (Rad-sig). To develop a clinical-radiological model, univariate and multivariate logistic regression analyses were performed. This model incorporated relevant clinical information, traditional imaging characteristics, and DLCT parameters to deliver a holistic evaluation. Finally, the Rad-sig from the best-performing radiomics model was combined with independent predictive factors from the clinical-radiological model to develop a nomogram. The entire study workflow is depicted in Figure 2 . Statistical analysis Statistical analyses were performed using Python (version 3.7.0), SPSS (version 26.0) and R (version 4.1.0) software. Continuous variables were assessed for normality with the Shapiro-Wilk test and for variance homogeneity with the Levene test. Normally distributed variables were analyzed using parametric tests (t-test or analysis of variance), while nonparametric alternatives (Mann-Whitney U test and Kruskal-Wallis test) were applied to non-normally distributed variables. Categorical data were analyzed using either the Chi-square or Fisher’s exact tests. For multiple comparisons, the Bonferroni correction method was used for adjustment. Statistical significance was defined as P <0.05. Receiver operating characteristic (ROC) curve analysis was performed to evaluate model performance in predicting STAS, with calculation of sensitivity, specificity, accuracy, and area under the curve (AUC). The DeLong test was used to compare AUC values between models. Calibration curves and the Hosmer-Lemeshow test assessed agreement between predicted probabilities and observed outcomes. Decision curve analysis (DCA) evaluated the clinical net benefit of the models. Results Patient characteristics A total of 266 patients diagnosed with LUAD (124 Men, 142 Women; ages 35–83 years) were divided into three groups: the training set (131 patients, 40 STAS-positive, 91 STAS-negative), the internal validation set (57 patients, 18 STAS-positive, 39 STAS-negative), and the external validation set (78 patients, 23 STAS-positive, 55 STAS-negative). Except for the lobulation sign ( P = 0.032), no significant differences were observed in clinical or conventional radiological features across the three cohorts. The baseline characteristics are shown in Table 1 . Table 1. Clinical characteristics and radiology parameters Training set (n=131) Internal Validation Set (n=57) External Validation Set (n=78) P value Age a 62(56.0, 69.0) 65(59.0, 71.0) 61(66.0, 69.0) 0.245 Gender 0.198 Man 68(51.91) 22(38.60) 34(43.59) Woman 63(48.09) 35(61.40) 44(56.41) Smoking history 0.323 No 85(64.89) 37(64.91) 58(74.36) YES 46(35.11) 20(35.09) 20(25.64) Location 0.671 Left upper lobe 29(22.14) 21(36.84) 25(32.05) Left lower lobe 20(15.27) 8(14.04) 12(15.38) Right upper lobe 41(31.30) 16(28.07) 21(26.92) Right middle lobe 14(10.69) 4(7.02) 6(7.69) Right lower lobe 27(20.61) 8(14.04) 14(17.95) LD 0.533 ≤ 20 mm 59(45.04) 30(52.63) 40(51.28) > 20 mm 72(54.96) 27(47.37) 38(48.72) CTR a 0.64(0.26, 1.00) 0.63(0.34, 1.00) 0.71(0.29, 1.00) 0.812 Nodule type 0.965 pGGN 27(20.61) 10(17.54) 13(16.67) Mixed solid 57(43.51) 26(45.62) 36(46.15) Solid nodule 47(35.88) 21(36.84) 29(37.18) Lobulation 0.032 b Absent 45(34.35) 14(24.56) 14(17.95) Present 86(65.65) 43(75.44) 64(82.05) Spiculation 0.071 Absent 85(64.89) 33(57.89) 38(48.72) Present 46(35.11) 24(42.11) 40(51.28) Shape 0.333 Round to oval 58(44.27) 31(54.39) 33(42.31) Irregular 73(55.73) 26(45.61) 45(57.69) Air bronchogram 0.472 Absent 76(58.02) 32(56.14) 51(65.38) Present 55(41.98) 25(43.86) 27(34.62) Tumor-lung interface 0.785 Well-defined 102(77.86) 45(78.95) 58(74.36) Ill-defined 29(22.14) 12(21.05) 20(25.64) Vacuole sign 0.948 Absent 103(78.63) 45(78.95) 60(76.92) Present 28(21.37) 12(21.05) 18(23.08) Pleural indentation 0.220 Absent 68(51.91) 22(38.60) 35(44.87) Present 63(48.09) 35(61.40) 43(55.13) λHU a 2.06(1.71,2.48) 2.00(1.59,2.53) 2.26(1.81,2.77) 0.135 NID 0.36±0.08 0.35±0.10 0.38±0.11 0.089 Zeff 8.52±0.35 8.48±0.37 8.54±0.33 0.616 ED a 89.50(67.85, 98.55) 88.00(72.10, 98.30) 92.40(71.40, 97.80) 0.950 a Data were presented as medians (25th-75th percentiles) and compared using the Kruskal-Wallis rank sum test. b Statistical significance was found between the training and external validation sets. LD long-axis diameter, CTR consolidation/tumor ratio, pGGN pure ground-glass nodule, λHU spectral curve slope, NID normalized iodine density, Zeff effective atomic number, ED electron density. Clinical-radiological model Univariate logistic regression analysis of the training dataset revealed several clinical and radiological factors associated with STAS ( P <0.05). These factors included tumor LD, CTR, nodule type, tumor-lung interface, pleural indentation, Zeff value, and ED value ( Table 2 ). Following multivariate analysis, tumor-lung interface (OR: 3.54; 95% CI: 1.09–11.46; P = 0.035) and ED value (OR: 1.17; 95% CI: 1.02–1.34; P = 0.028) were identified as independent predictors of STAS. The clinical-radiological model demonstrated satisfactory predictive performance, with AUCs of 0.870 (95% CI: 0.805–0.935) in the training set, 0.798 (95% CI: 0.675–0.922) in the internal validation set, and 0.819 (95% CI: 0.717–0.921) in the external validation set. Table 2. Univariate and multivariate logistic regression analyses of clinical-radiological characteristics in training set Variables Univariate analysis Multivariate analysis OR (95%CI) P value OR (95%CI) P value Age 1.02 (0.98 ~ 1.06) 0.363 Gender 1.72 (0.81 ~ 3.66) 0.155 Smoking history 1.58 (0.73 ~ 3.40) 0.242 Location Left upper lobe 1.00 (Reference) Left lower lobe 2.10 (0.61 ~ 7.20) 0.24 Right upper lobe 1.46 (0.50 ~ 4.28) 0.491 Right middle lobe 2.36 (0.61 ~ 9.16) 0.216 Right lower lobe 0.90 (0.26 ~ 3.11) 0.865 LD 3.50 (1.53 ~ 7.99) 0.003 * 1.54 (0.53 ~ 4.46) 0.431 CTR 142.77(18.92 ~ 1077.15) <0.001 * 1.93 (0.02 ~ 242.39) 0.791 Nodule type pGGN 1.00 (Reference) 1.00 (Reference) Mixed solid 5.53 (0.67 ~ 45.62) 0.112 0.02 (0.00 ~ 1.46) 0.072 Solid nodule 41.89 (5.23 ~ 335.72) <0.001 * 0.01 (0.00 ~ 4.01) 0.139 Lobulation 1.87 (0.82 ~ 4.30) 0.138 Spiculation 1.16 (0.54 ~ 2.52) 0.705 Shape 2.03 (0.93 ~ 4.43) 0.074 Air bronchogram 0.56 (0.26 ~ 1.22) 0.147 Tumor-lung interface 3.30 (1.40 ~ 7.77) 0.006 * 3.54 (1.09 ~ 11.46) 0.035 * Vacuole sign 1.65 (0.69 ~ 3.95) 0.259 Pleural indentation 2.71 (1.25 ~ 5.87) 0.011 * 0.74 (0.25 ~ 2.20) 0.594 λHU 0.66 (0.35 ~ 1.25) 0.203 NID 6.48 (0.07 ~ 591.82) 0.417 Zeff 0.05 (0.01 ~ 0.21) <0.001 * 0.61 (0.09 ~ 4.24) 0.620 ED 1.13 (1.07 ~ 1.19) <0.001 * 1.17 (1.02 ~ 1.34) 0.028 * LD long-axis diameter, CTR consolidation/tumor ratio, pGGN pure ground-glass nodule, λHU spectral curve slope, NID normalized iodine density, Zeff effective atomic number, ED electron density. * p < 0.05 Feature selection and performance of radiomics model A total of 1,834 radiomic features were extracted from seven imaging modalities: Conventional, VMI 40keV, 65keV, 100keV, ID, Zeff, and ED. LASSO regression identified 8-19 features with non-zero coefficients for multi-parametric DLCT imaging, and their feature-weight distributions are depicted in Supplemental Figure 2 . An LR classifier was then employed to construct the radiomics model. Figure 3 illustrates ROC curves for the seven models, and Table 3 summarizes their predictive metrics. The models exhibited AUC ranges of 0.847–0.945 (training), 0.751–0.835 (internal validation), and 0.768–0.828 (external validation). To identify the most suitable model for diagnostic development, we compared the performance of different sequences within the radiomics models. Among these, the VMI 40keV-based radiomics model demonstrated the best predictive performance, achieving an AUC of 0.899 (95% CI: 0.843–0.955) in the training set, 0.835 (95% CI: 0.711–0.958) in the internal validation set, and 0.828 (95% CI: 0.736–0.921) in the external validation set. Given its consistently superior performance across all datasets, the VMI 40keV model was selected as the final model for nomogram construction. Table 3. Performance of different radiomics models in predicting STAS status of lung adenocarcinoma in the training set, internal validation set, and external validation set Model AUC (95% CI) Accuracy Sensitivity Specificity Training set Conventional 0.885 (0.830 - 0.941) 0.763 0.900 0.703 VMI 40keV 0.899 (0.843 - 0.955) 0.802 0.825 0.791 VMI 65keV 0.894 (0.839 - 0.948) 0.832 0.850 0.824 VMI 100keV 0.910 (0.861 - 0.959) 0.847 0.800 0.868 Iodine density 0.945 (0.911 - 0.980) 0.870 0.950 0.835 Zeff 0.847 (0.773 - 0.921) 0.824 0.650 0.901 Electron density 0.890 (0.834 - 0.947) 0.786 0.900 0.736 Internal validation set Conventional 0.775 (0.646 - 0.904) 0.719 0.722 0.718 VMI 40keV 0.835 (0.711 - 0.958) 0.789 0.833 0.769 VMI 65keV 0.793 (0.663 - 0.924) 0.702 0.833 0.641 VMI 100keV 0.799 (0.673 - 0.925) 0.719 0.778 0.692 Iodine density 0.751 (0.610 - 0.891) 0.772 0.389 0.949 Zeff 0.825 (0.701 - 0.948) 0.772 0.833 0.744 Electron density 0.812 (0.682 - 0.942) 0.754 0.667 0.795 External validation set Conventional 0.801 (0.704 - 0.898) 0.705 0.826 0.655 VMI 40keV 0.828 (0.736 - 0.921) 0.782 0.739 0.800 VMI 65keV 0.804 (0.703 - 0.904) 0.718 0.783 0.691 VMI 100keV 0.774 (0.665 - 0.883) 0.718 0.696 0.727 Iodine density 0.795 (0.694 - 0.896) 0.756 0.783 0.745 Zeff 0.769 (0.655 - 0.884) 0.756 0.783 0.745 Electron density 0.794 (0.690 - 0.899) 0.756 0.826 0.727 STAS spread through air spaces, AUC area under the curve, CI c onfidence interval, VMI virtual monoenergetic images, Zeff effective atomic number. Nomogram development and validation A nomogram ( Figure 4A ) was developed by integrating independent predictors (tumor-lung interface and ED value) with the Rad-sig derived from the optimal radiomics model (VMI 40keV). Compared with the clinical-radiological model, the nomogram achieved higher AUCs for distinguishing STAS (0.910 vs 0.870, training set; 0.868 vs 0.798, internal validation set; 0.848 vs 0.819, external validation set). In the training and internal validation sets, the nomogram showed significantly higher performance than the clinical-radiological model ( P < 0.05), whereas no significant improvement was observed in the external validation set ( P = 0.184). Furthermore, the performance of the nomogram did not differ significantly from that of the radiomics model across all three datasets ( P > 0.05). Table 4 summarizes the predictive performance of each model, with the associated ROC curves shown in Figure 4B-D . Results of the DeLong test for each model are provided in Supplemental Figure 3 . As shown in Figure 5A-C , the nomogram demonstrated satisfactory calibration across the three cohorts. The Hosmer–Lemeshow test indicated that the nomogram predictions (training set: P = 0.383; internal validation set: P = 0.125; external validation set: P = 0.083) closely aligned with the actual outcomes, with no significant difference. Decision curve analysis further revealed that, across all datasets, the net benefits of the clinical-radiological, radiomics, and nomogram models generally surpassed those of either treating all patients or not across most reasonable threshold ranges. Additionally, when the risk threshold was less than 0.5, the clinical-radiological model demonstrated a relatively lower overall net benefit ( Figure 5D-F ). Table 4. Predictive performance of clinical-radiologic model, radiomics model, and nomogram Model AUC (95% CI) Accuracy Sensitivity Specificity Training set Clinical-radiological 0.870 (0.805 - 0.935) 0.779 0.900 0.725 VMI 40KeV 0.899 (0.843 - 0.955) 0.802 0.825 0.791 Nomogram 0.910 (0.856 - 0.965) 0.855 0.875 0.846 Internal validation set Clinical-radiological 0.798 (0.675 - 0.922) 0.702 0.778 0.667 VMI 40KeV 0.835 (0.711 - 0.958) 0.789 0.833 0.769 Nomogram 0.868 (0.757 - 0.978) 0.842 0.778 0.872 External validation set Clinical-radiological 0.819 (0.717 - 0.921) 0.756 0.783 0.745 VMI 40KeV 0.828 (0.736 - 0.921) 0.782 0.739 0.800 Nomogram 0.848 (0.752 - 0.944) 0.795 0.826 0.782 AUC area under the curve, CI c onfidence interval, VMI virtual monoenergetic images. Discussion Improving non-invasive and objective methods for distinguishing aggressive subtypes of LUAD, such as STAS, holds significant clinical value. This study evaluated the diagnostic performance of multi-parametric DLCT for identifying STAS. In this dual-center study, we found that ED values and tumor-lung interface were independent predictors for identifying STAS. The nomogram that incorporated these two predictive factors, along with the VMI 40keV radiomics signature , outperformed the clinical-radiological model, providing superior diagnostic accuracy of STAS status in LUAD. The incidence of STAS in LUAD has been found to vary widely across different studies, with reported rates ranging from 15.7% to 58.4% [13,22]. In this study, the overall STAS incidence was 30.5%, consistent with previously published studies. Previous studies have shown that certain CT morphological characteristics, such as nodule type, spiculation, air bronchogram, tumor-lung interface and pleural indentation, are related to STAS status[2,13,23,24]. The results of this study are generally consistent with these findings, but there are slight differences in spiculation and air bronchogram. The analysis of quantitative parameters, revealed that STAS was significantly related to both increased tumor LD and increased CTR. The work of Kim et al[2] and Tasnim et al[25] also supports this result. However, when ED was incorporated into the multivariate analysis, the results indicated that ED was an independent predictor of STAS, whereas CTR was excluded. We speculate that ED, as a DLCT quantitative parameter, may reflect the physical density of the tumor, thereby providing insights into its pathological characteristics[26]. Tumors with a high proportion of solid components typically present higher CT values, which could explain the observed correlation between these characteristics. Therefore, we hypothesize that ED can characterize STAS more effectively than CTR. A study conducted by Liu et al[17] supported this conclusion, as their model, which utilized LD and venous-phase ED values in 225 patients with LUAD, demonstrated satisfactory performance (AUC = 0.840). However, the absence of an independent validation set limits the assessment of its external applicability. In contrast, the clinical-radiological model in our study, which incorporates tumor-lung interface and ED values, exhibited robust applicability and reliability across diverse datasets. Most radiomics-related studies have extracted features from conventional CT and PET/CT images to build models[27,28]. Recently, studies have shifted their focus to the importance of DLCT images in radiomics, suggesting that DLCT image-derived features could enhance model performance[29,30]. Radiomic features were extracted from a variety of imaging modalities in this study, including conventional images, VMI at 40keV, 65keV, and 100keV, as well as ID, Zeff, and ED maps. Consequently, seven distinct radiomics models were developed. Among these, the model based on VMI 40keV demonstrated the best performance. Several studies have confirmed that, compared to conventional reconstructed images, VMI 40keV provides superior tissue contrast, thereby enhancing the accuracy of lesion detection[31,32]. A previous DLCT imaging study on thoracic tumors demonstrated that VMI in the energy range of 40 to 70keV have a significantly higher signal-to-noise ratio (SNR) compared to conventional reconstructed images, with the SNR progressively improving as the energy level decreases[33]. In terms of subjective visual assessment, VMI at 40keV outperformed both 70keV and conventional reconstructed images[33]. Wang et al found that among VMI 40-140keV (with intervals of 10keV), the deep learning model based on VMI 40keV outperformed other energy levels in predicting lymph node metastasis in lung cancer[34], which is consistent with our findings. This suggests that imaging at lower energy levels may provide more detailed information about tumor angiogenesis and heterogeneity[34]. These results provide a theoretical foundation for the superiority of the radiomics model based on VMI 40keV compared to other images. Additionally, we developed a nomogram that integrates conventional imaging features, DLCT quantitative parameters, and radiomics signature from VMI 40keV to predict STAS. This nomogram achieved optimal predictive performance, significantly outperforming the clinical-radiological model in both the training and internal validation sets ( P 0.05), the nomogram still achieved higher AUC, accuracy, sensitivity, and specificity than the clinical-radiological model. This enhanced performance may stem from the nomogram’s ability to reveal potential heterogeneity in LUAD across multiple dimensions[35]. This study has certain limitations. First, the relatively small sample size of this retrospective study may introduce potential selection and statistical bias. Second, images were obtained exclusively using the Philips Spectral CT scanner, so the predictive performance and reproducibility of the model with DLCT images acquired from other manufacturers require further validation. Finally, combining radiomics models from both tumor and peritumoral regions may enhance the nomogram’s predictive performance. To validate the predictive significance of peritumoral information, large-scale, multicenter prospective studies are essential. In conclusion, w e constructed and validated a DLCT-based nomogram, incorporating conventional radiological features, DLCT quantitative parameters, and VMI 40keV radiomic features. This method represents an effective, non-invasive, and feasible auxiliary diagnostic tool that aids clinicians in preoperative prediction of STAS, supporting clinical decision-making for patients with LUAD. Abbreviations DLCT: Dual-layer spectral detector CT STAS: Spread through air spaces LUAD: Lung adenocarcinoma VMI: Virtual monoenergetic images ID: Iodine density Zeff: Effective atomic number ED: Electron density SBI: Spectral-based images LD: Long-axis diameter CTR: Consolidation/tumor ratio ROI: Region of interest λHU: Spectral curve slope ICC: Inter-class correlation coefficient LASSO: Least absolute shrinkage and selection operator LR: Logistic regression AUC: Area under the curve DCA: Decision curve analysis Declarations Acknowledgement We sincerely thank all the individuals who participated in these studies, as well as the researchers and technicians whose contributions made this work possible. Funding This study was supported by the National Natural Science Foundation of China (grant numbers:82160335, 82202140) and the National Key R&D Program of China (grant number: 2022YFC2010002). Author information Pei Huang, Ze Lin and Yingying Qiu contributed equally to this article. Authors and Affiliations Department of Radiology, Jiangxi Provincial People’s Hospital, The First Affiliated Hospital of Nanchang Medical College, Nanchang 330006, China Pei Huang, Yingying Qiu, Dan Chen & Bing Fan Department of Radiology, Wuhan Third Hospital, Tongren Hospital of Wuhan University, Wuhan, Hubei , China Ze Lin & Yiqing Tan Department of Radiology, the Second Affiliated Hospital of Naval Medical University, Shanghai 200003, China Li Fan Department of Radiology, the Affiliated Hospital of Guizhou Medical University, Guiyang 550004, China Pinggui Lei Clinical & Technical Support, Philips Healthcare, Shanghai, People’s Republic of China Yuting Liao Department of Radiology, The Second Affiliated Hospital, Jiangxi Medical College, Nanchang University, Nanchang 330006, China Jiangxi Provincial Key Laboratory of Intelligent Medical Imaging, Nanchang 330006, China Minjing Zuo Contributions P.H., Z.L., Y.Q. and B.F. conceptualized the study. P.H., Z.L., D.C. and Y.T. organized the database. P.H. and Z.L. wrote the manuscript. P.H., Z.L. and Y.L. performed the statistical analyses. L.F., P.L. and M.Z. provided critical feedback and discussions. B.F. edited the manuscript. M.Z. and B.F. supervised this study. All authors have read and approved the final manuscript. Corresponding authors Correspondence to Minjing Zuo or Bing Fan. Ethics approval and consent to participate The study was approved by the Ethics Committee of Jiangxi Provincial People’s Hospital (ethics approval number: KK-2024-029) and the Second Affiliated Hospital of Nanchang University (ethics approval number: IIT-O-2024-066), following the Declaration of Helsinki. Due to the retrospective nature of the research, the Ethics Committee waived the requirement for informed consent. Consent for publication Not applicable. Competing interest Yuting Liao is employee of Philips Healthcare (Shanghai, China). The remaining authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article. References Travis WD, Brambilla E, Nicholson AG, Yatabe Y, Austin JHM, Beasley MB, et al. The 2015 World Health Organization Classification of Lung Tumors: Impact of Genetic, Clinical and Radiologic Advances Since the 2004 Classification. Journal of Thoracic Oncology. 2015; 10:1243-1260. Kim SK, Kim TJ, Chung MJ, Kim TS, Lee KS, Zo JI, et al. Lung Adenocarcinoma: CT Features Associated with Spread through Air Spaces. Radiology. 2018; 289:831-840. Toyokawa G, Yamada Y, Tagawa T, Oda Y. Significance of spread through air spaces in early-stage lung adenocarcinomas undergoing limited resection. Thoracic Cancer. 2018; 9:1255-1261. Blaauwgeers H, Flieder D, Warth A, Harms A, Monkhorst K, Witte B, et al. A Prospective Study of Loose Tissue Fragments in Non-Small Cell Lung Cancer Resection Specimens: An Alternative View to "Spread Through Air Spaces". American Journal of Surgical Pathology. 2017; 41:1226-1230. Warth A, Muley T, Kossakowski CA, Goeppert B, Schirmacher P, Dienemann H, et al. Prognostic Impact of Intra-alveolar Tumor Spread in Pulmonary Adenocarcinoma. American Journal of Surgical Pathology. 2015; 39:793-801. Aly RG, Rekhtman N, Li X, Takahashi Y, Eguchi T, Tan KS, et al. Spread Through Air Spaces (STAS) Is Prognostic in Atypical Carcinoid, Large Cell Neuroendocrine Carcinoma, and Small Cell Carcinoma of the Lung. Journal of Thoracic Oncology. 2019; 14:1583-1593. Laville D, Desage AL, Fournel P, Bayle-Bleuez S, Neifer C, Picot T, et al. Spread Through Air Spaces in Stage I to III Resected Lung Adenocarcinomas: Should the Presence of Spread Through Air Spaces Lead to an Upstaging? American Journal of Surgical Pathology. 2024; 48:596-604. Eguchi T, Kameda K, Lu S, Bott MJ, Tan KS, Montecalvo J, et al. Lobectomy Is Associated with Better Outcomes than Sublobar Resection in Spread through Air Spaces (STAS)-Positive T1 Lung Adenocarcinoma: A Propensity Score-Matched Analysis. Journal of Thoracic Oncology. 2019; 14:87-98. Altorki N, Wang X, Kozono D, Watt C, Landrenau R, Wigle D, et al. Lobar or Sublobar Resection for Peripheral Stage IA Non-Small-Cell Lung Cancer. N Engl J Med. 2023; 388:489-498. Loganadane G, Martinetti F, Mercier O, Krhili S, Riet FG, Mbagui R, et al. Stereotactic ablative radiotherapy for early stage non-small cell lung cancer: A critical literature review of predictive factors of relapse. Cancer Treatment Reviews. 2016; 50:240-246. Villalba JA, Shih AR, Sayo T, Kunitoki K, Hung YP, Ly A, et al. Accuracy and Reproducibility of Intraoperative Assessment on Tumor Spread Through Air Spaces in Stage 1 Lung Adenocarcinomas. Journal of Thoracic Oncology. 2021; 16:619-629. Zhou F, Villalba JA, Sayo T, Narula N, Pass H, Mino-Kenudson M, et al. Assessment of the feasibility of frozen sections for the detection of spread through air spaces (STAS) in pulmonary adenocarcinoma. Mod Pathol. 2022; 35:210-217. Toyokawa G, Yamada Y, Tagawa T, Kamitani T, Yamasaki Y, Shimokawa M, et al. Computed tomography features of resected lung adenocarcinomas with spread through air spaces. J Thorac Cardiovasc Surg. 2018; 156:1670-1676. Moore J, Remy J, Altschul E, Chusid J, Flohr T, Raoof S, et al. Thoracic Applications of Spectral CT Scan. Chest. 2024; 165:417-430. Agostini A, Borgheresi A, Mari A, Floridi C, Bruno F, Carotti M, et al. Dual-energy CT: theoretical principles and clinical applications. Radiologia Medica. 2019; 124:1281-1295. Ha T, Kim W, Cha J, Lee YH, Seo HS, Park SY, et al. Differentiating pulmonary metastasis from benign lung nodules in thyroid cancer patients using dual-energy CT parameters. European Radiology. 2022; 32:1902-1911. Liu BC, Ma HY, Huang J, Luo YW, Zhang WB, Deng WW, et al. Does dual-layer spectral detector CT provide added value in predicting spread through air spaces in lung adenocarcinoma? A preliminary study. European Radiology. 2024; 34:4176-4186. Wu F, Zhou H, Li F, Wang JT, Ai T. Spectral CT Imaging of Lung Cancer: Quantitative Analysis of Spectral Parameters and Their Correlation with Tumor Characteristics. Academic Radiology. 2018; 25:1398-1404. Mayerhoefer ME, Materka A, Langs G, Haggstrom I, Szczypinski P, Gibbs P, et al. Introduction to Radiomics. Journal of Nuclear Medicine. 2020; 61:488-495. Perez-Johnston R, Araujo-Filho JA, Connolly JG, Caso R, Whiting K, Tan KS, et al. CT-based Radiogenomic Analysis of Clinical Stage I Lung Adenocarcinoma with Histopathologic Features and Oncologic Outcomes. Radiology. 2022; 303:664-672. Choi W, Oh JH, Riyahi S, Liu CJ, Jiang F, Chen W, et al. Radiomics analysis of pulmonary nodules in low-dose CT for early detection of lung cancer. Medical Physics. 2018; 45:1537-1549. Shiono S, Endo M, Suzuki K, Yanagawa N. Spread through air spaces affects survival and recurrence of patients with clinical stage IA non-small cell lung cancer after wedge resection. Journal of Thoracic Disease. 2020; 12:2247-2260. Qi L, Xue K, Cai Y, Lu J, Li X, Li M. Predictors of CT Morphologic Features to Identify Spread Through Air Spaces Preoperatively in Small-Sized Lung Adenocarcinoma. Frontiers in Oncology. 2020; 10:548430. Zhang X, Qiao W, Shen J, Jiang Q, Pan C, Wang Y, et al. Clinical, pathological, and computed tomography morphological features of lung cancer with spread through air spaces. Transl Lung Cancer Res. 2024; 13:2802-2812. Tasnim S, Raja S, Mukhopadhyay S, Blackstone EH, Toth AJ, Barron JO, et al. Preoperative predictors of spread through air spaces in lung cancer. J Thorac Cardiovasc Surg. 2024; 168:660-669. Tatsugami F, Higaki T, Kiguchi M, Tsushima S, Taniguchi A, Kaichi Y, et al. Measurement of electron density and effective atomic number by dual-energy scan using a 320-detector computed tomography scanner with raw data-based analysis: a phantom study. J Comput Assist Tomogr. 2014; 38:824-827. Guo R, Yan W, Wang F, Su H, Meng X, Xie Q, et al. The utility of (18)F-FDG PET/CT for predicting the pathological response and prognosis to neoadjuvant immunochemotherapy in resectable non-small-cell lung cancer. Cancer Imaging. 2024; 24:120. Dong Y, Jiang Z, Li C, Dong S, Zhang S, Lv Y, et al. Development and validation of novel radiomics-based nomograms for the prediction of EGFR mutations and Ki-67 proliferation index in non-small cell lung cancer. Quant Imaging Med Surg. 2022; 12:2658-2671. Li M, Fan Y, You H, Li C, Luo M, Zhou J, et al. Dual-Energy CT Deep Learning Radiomics to Predict Macrotrabecular-Massive Hepatocellular Carcinoma. Radiology. 2023; 308:e230255. Xu H, Zhu N, Yue Y, Guo Y, Wen Q, Gao L, et al. Spectral CT-based radiomics signature for distinguishing malignant pulmonary nodules from benign. Bmc Cancer. 2023; 23:91. Nagayama Y, Tanoue S, Inoue T, Oda S, Nakaura T, Utsunomiya D, et al. Dual-layer spectral CT improves image quality of multiphasic pancreas CT in patients with pancreatic ductal adenocarcinoma. European Radiology. 2020; 30:394-403. Wang X, Liu D, Jiang S, Zeng X, Li L, Yu T, et al. Subjective and Objective Assessment of Monoenergetic and Polyenergetic Images Acquired by Dual-Energy CT in Breast Cancer. Korean Journal of Radiology. 2021; 22:502-512. Doerner J, Hauger M, Hickethier T, Byrtus J, Wybranski C, Grosse HN, et al. Image quality evaluation of dual-layer spectral detector CT of the chest and comparison with conventional CT imaging. European Journal of Radiology. 2017; 93:52-58. Wang YW, Chen CJ, Wang TC, Huang HC, Chen HM, Shih JY, et al. Multi-energy level fusion for nodal metastasis classification of primary lung tumor on dual energy CT using deep learning. Computers in Biology and Medicine. 2022; 141:105185. Zhu M, Yang Z, Wang M, Zhao W, Zhu Q, Shi W, et al. A computerized tomography-based radiomic model for assessing the invasiveness of lung adenocarcinoma manifesting as ground-glass opacity nodules. Respir Res. 2022; 23:96. Additional Declarations Competing interest reported. Yuting Liao is employee of Philips Healthcare (Shanghai, China). The remaining authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article. Supplementary Files SupplementaryMaterial.docx Cite Share Download PDF Status: Published Journal Publication published 22 Dec, 2025 Read the published version in BMC Cancer → Version 1 posted Editorial decision: Revision requested 03 Oct, 2025 Reviews received at journal 01 Oct, 2025 Reviewers agreed at journal 26 Sep, 2025 Reviews received at journal 21 Sep, 2025 Reviewers agreed at journal 21 Sep, 2025 Reviewers invited by journal 09 Sep, 2025 Editor assigned by journal 18 Aug, 2025 Submission checks completed at journal 18 Aug, 2025 First submitted to journal 17 Aug, 2025 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-7391071","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":512561910,"identity":"9e695cb9-b0b8-4db3-8f46-99318cfab5d5","order_by":0,"name":"Pei Huang","email":"","orcid":"","institution":"The First Affiliated Hospital of Nanchang Medical College","correspondingAuthor":false,"prefix":"","firstName":"Pei","middleName":"","lastName":"Huang","suffix":""},{"id":512561911,"identity":"ac89cd95-9c7f-4caf-ac34-92ee40409589","order_by":1,"name":"Ze Lin","email":"","orcid":"","institution":"Wuhan Third Hospital, Tongren Hospital of Wuhan University","correspondingAuthor":false,"prefix":"","firstName":"Ze","middleName":"","lastName":"Lin","suffix":""},{"id":512561912,"identity":"179019a6-5886-4f51-8018-650be33cfffd","order_by":2,"name":"Yingying Qiu","email":"","orcid":"","institution":"The First Affiliated Hospital of Nanchang Medical College","correspondingAuthor":false,"prefix":"","firstName":"Yingying","middleName":"","lastName":"Qiu","suffix":""},{"id":512561913,"identity":"5d1bf665-22dc-48b3-8df6-647ea88dd212","order_by":3,"name":"Dan Chen","email":"","orcid":"","institution":"The First Affiliated Hospital of Nanchang Medical College","correspondingAuthor":false,"prefix":"","firstName":"Dan","middleName":"","lastName":"Chen","suffix":""},{"id":512561914,"identity":"e2a9d503-fb46-48b6-a994-8bf33ba0fc82","order_by":4,"name":"Yiqing Tan","email":"","orcid":"","institution":"Wuhan Third Hospital, Tongren Hospital of Wuhan University","correspondingAuthor":false,"prefix":"","firstName":"Yiqing","middleName":"","lastName":"Tan","suffix":""},{"id":512561915,"identity":"b3ceb81f-3509-4b99-a95f-98b884e8edd3","order_by":5,"name":"Li Fan","email":"","orcid":"","institution":"the Second Affiliated Hospital of Naval Medical University","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Fan","suffix":""},{"id":512561916,"identity":"69014137-04e4-4931-8be1-b3fa552e3383","order_by":6,"name":"Pinggui Lei","email":"","orcid":"","institution":"the Affiliated Hospital of Guizhou Medical University","correspondingAuthor":false,"prefix":"","firstName":"Pinggui","middleName":"","lastName":"Lei","suffix":""},{"id":512561917,"identity":"dcfaa123-06e8-4a68-97af-267cc05c007d","order_by":7,"name":"Yuting Liao","email":"","orcid":"","institution":"Clinical \u0026 Technical Support, Philips Healthcare","correspondingAuthor":false,"prefix":"","firstName":"Yuting","middleName":"","lastName":"Liao","suffix":""},{"id":512561918,"identity":"6da510ca-46d6-4c3b-8954-678a0d246e0f","order_by":8,"name":"Minjing Zuo","email":"","orcid":"","institution":"Department of Radiology, The Second Affiliated Hospital, Jiangxi Medical College, Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Minjing","middleName":"","lastName":"Zuo","suffix":""},{"id":512561919,"identity":"7bf4f511-e08b-4685-9615-433c1ab96c47","order_by":9,"name":"Bing Fan","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvUlEQVRIiWNgGAWjYBACfvbG9h8JP2x4+InWItlzuEHiYU+ajGQDsVoMbrg3SD5gO2xjcIBoa24wNhgk8KTxGB9P3sDwo2IbYR2MsxsbEhIsbHjMzjwrYOw5c5uwFmaZgw0HQLaY3cgxYGZsI0ILm0Qi0Bq2wzzGM4jVwiOR2MwA0mIgQawWCZ6DbQyJPWk8EkC/HCTKL/bH258x/vhhY8/fnrzxwY8KIrQggQQSogauhVQdo2AUjIJRMEIAABwVPq7SvO+XAAAAAElFTkSuQmCC","orcid":"","institution":"The First Affiliated Hospital of Nanchang Medical College","correspondingAuthor":true,"prefix":"","firstName":"Bing","middleName":"","lastName":"Fan","suffix":""}],"badges":[],"createdAt":"2025-08-17 07:53:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7391071/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7391071/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12885-025-15436-7","type":"published","date":"2025-12-22T15:57:31+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":91509858,"identity":"e9c2b249-e873-4aa2-bc58-6e57a31725bf","added_by":"auto","created_at":"2025-09-17 08:41:34","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":120496,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart for patient selection. STAS spread through air spaces, DLCT dual-layer spectral detector CT, AIS adenocarcinoma in situ, MIA minimally invasive adenocarcinoma, SBI spectral-based images.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7391071/v1/466258188a6b6b7b8e58bee4.png"},{"id":91509859,"identity":"b9dc247e-143c-4ec6-b619-f77956da6434","added_by":"auto","created_at":"2025-09-17 08:41:34","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":303293,"visible":true,"origin":"","legend":"\u003cp\u003eOverall workflow of this study. ROI region of interest, λHU spectral curve slope, NID normalized iodine density, Zeff effective atomic number, ED electron density, LD long-axis diameter, CTR consolidation/tumor ratio, LASSO least absolute shrinkage and selection operator, ROC receiver operating characteristic, DCA decision curve analysis.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7391071/v1/fcda89fc98df0555cfc4ee0e.png"},{"id":91511430,"identity":"860c57ac-d4cf-449e-affb-2bf8516c80cc","added_by":"auto","created_at":"2025-09-17 08:49:34","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":145066,"visible":true,"origin":"","legend":"\u003cp\u003eReceiver operating characteristic\u003cstrong\u003e \u003c/strong\u003ecurves\u003cstrong\u003e \u003c/strong\u003eof different radiomics models in the training set (\u003cstrong\u003eA\u003c/strong\u003e), internal validation set (\u003cstrong\u003eB\u003c/strong\u003e), and external validation set (\u003cstrong\u003eC\u003c/strong\u003e). VMI virtual monoenergetic images, ID iodine density, Zeff effective atomic number, ED electron density.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7391071/v1/c4249933125adf4daf9abf36.png"},{"id":91509862,"identity":"6ba0b8b2-6b0d-4420-b9b8-5fbebf54e984","added_by":"auto","created_at":"2025-09-17 08:41:34","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":162187,"visible":true,"origin":"","legend":"\u003cp\u003eNomogram for predicting spread through air spaces (STAS) of lung adenocarcinoma developed based on the training set (\u003cstrong\u003eA\u003c/strong\u003e). Receiver operating characteristic\u003cstrong\u003e \u003c/strong\u003ecurves of clinical-radiological model, VMI 40keV-based radiomics model, and nomogram in the training, internal validation, and external validation sets (\u003cstrong\u003eB-D\u003c/strong\u003e). ED electron density, Rad-Sig radiomics signature, VMI virtual monoenergetic images.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7391071/v1/5ba125728c773428efb63697.png"},{"id":91511431,"identity":"176e8f3c-443d-4df2-9ba9-f4b662c08756","added_by":"auto","created_at":"2025-09-17 08:49:34","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":198754,"visible":true,"origin":"","legend":"\u003cp\u003eCalibration\u003cstrong\u003e \u003c/strong\u003ecurves (\u003cstrong\u003eA-C\u003c/strong\u003e) and Decision curves analysis (\u003cstrong\u003eD-F\u003c/strong\u003e) of clinical-radiological model, VMI 40keV-based radiomics model, and nomogram in the training , internal validation, and external validation sets. VMI virtual monoenergetic images.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7391071/v1/f6fee66c83777a2af99ffcce.png"},{"id":99172442,"identity":"d7850096-a078-4cc0-a51a-b63cf0ad002e","added_by":"auto","created_at":"2025-12-29 16:09:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2311320,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7391071/v1/bb38e6a6-573d-469a-a54a-40297ca69a4e.pdf"},{"id":91509880,"identity":"15cffdd9-d75a-4d2a-ba57-06532dc71603","added_by":"auto","created_at":"2025-09-17 08:41:35","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":13755595,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7391071/v1/a61ec5f1ff45a6fe479ba113.docx"}],"financialInterests":"Competing interest reported. Yuting Liao is employee of Philips Healthcare (Shanghai, China). The remaining authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.","formattedTitle":"Dual-layer spectral detector CT quantitative parameters and radiomics for predicting spread through air spaces of lung adenocarcinoma: A dual-center study","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe 2015 World Health Organization classification introduced tumor spread through air spaces (STAS) as a novel lung cancer dissemination pattern, defined as micropapillary clusters, solid nests, or single tumor cells in air spaces beyond the tumor boundary[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. This indicates that lung cancer can exhibit microscopic spread via air spaces. Current research on STAS predominantly focuses on lung adenocarcinoma (LUAD), the most common histologic subtype of lung cancer[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Several studies have suggested that certain STAS cases may be false positives caused by artifacts during surgery or pathological specimen preparation[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. However, substantial evidence confirms STAS as an independent prognostic factor associated with reduced survival and increased recurrence[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. For example, a previous study demonstrated that lobectomy yields superior survival outcomes compared with sublobar resection in stage I lung adenocarcinoma with STAS[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. In contrast, sublobar resection remains appropriate for early-stage patients with non-invasive histologic features, as it preserves pulmonary parenchyma and improves postoperative lung function and quality of life[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Furthermore, accurate STAS identification is essential for determining optimal margins in patients undergoing stereotactic radiotherapy, which critically impacts local recurrence [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e\u003cp\u003ePostoperative histopathological diagnosis is the gold standard for confirming STAS. Currently, there is no reliable method for detecting STAS through preoperative biopsy or intraoperative frozen section[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], which may limit its substantial impact on treatment decisions. Therefore, the development of a non-invasive, simple, and practical method to accurately predict STAS preoperatively could provide crucial information to guide clinical decision-making.\u003c/p\u003e\u003cp\u003eIn clinical practice, owing to its superior resolution and cost-effectiveness, high-resolution CT has become the preferred preoperative imaging method for pulmonary lesions. Previous research has demonstrated a correlation between some conventional imaging characteristics of LUAD and the presence of STAS[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e], but these features are largely dependent on the subjective experience of the interpreter, which may lead to potential bias. Recent advances in dual-layer spectral detector CT (DLCT) have transformed CT diagnosis from qualitative to quantitative assessment[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Beyond conventional images, DLCT produces energy-specific data such as virtual monoenergetic images (VMI), electron density (ED), and iodine density (ID) maps, enhancing diagnostic accuracy and clinical utility[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Many studies have shown that quantitative parameters from these images provide more accurate and reliable evidence for the differentiation of pulmonary lesions[\u003cspan additionalcitationids=\"CR17\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Additionally, radiomics enables the non-invasive assessment of tumor heterogeneity by rapidly extracting quantitative features from medical images[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. These features have demonstrated substantial efficacy in differentiating benign from malignant pulmonary lesions, categorizing molecular subtypes, and evaluating prognostic[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. However, the majority of radiomics research has focused on conventional CT imaging, with the potential usefulness of radiomic features from DLCT images remaining under-explored.\u003c/p\u003e\u003cp\u003eTo the best of our knowledge, few studies have investigated whether integrating these methods could provide additional benefits for evaluating pulmonary lesions. Therefore, this study aims to develop a nomogram that integrates clinical data, conventional imaging characteristics, DLCT quantitative parameters, and radiomic features for preoperative prediction of STAS in LUAD.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cp\u003e\u003cstrong\u003ePatients\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis retrospective study was conducted with the approval of the institutional ethics review board, which granted a waiver for written informed consent. Between September 2023 and July 2024, patients who underwent thoracic contrast-enhanced DLCT scans at Center 1 were included. After applying the specified inclusion and exclusion criteria, the eligible patients were randomly divided into training and internal validation sets at a 7:3 ratio. Similarly, between January 2024 and April 2024, patients from Center 2 who satisfied the same criteria were enrolled to establish the external validation set. Inclusion criteria: (1) postoperative pathology confirmed LUAD, and the status of STAS was determined; (2) initial diagnosis with no prior history of anti-tumor treatment; (3) DLCT scan performed within two weeks before surgery. Exclusion criteria: (1) postoperative pathological diagnosis of adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), or other special types such as mucinous, fetal, or intestinal-type adenocarcinoma; (2) images exhibiting severe artifacts, atelectasis, or obstructive pneumonia that obscures tumor margins; (3) incomplete clinical or pathological data; and (4) missing spectral-based images (SBI) data. The inclusion and exclusion flowchart for patient selection is shown in \u003cstrong\u003eFigure 1\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDLCT scan protocol\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDLCT scans were performed using a 256-slice Hawk Spectral CT at Center 1 and a 64-slice IQon Spectral CT at Center 2 (both Philips Healthcare, Netherlands).\u0026nbsp;\u0026nbsp;Detailed information on the DLCT scanning protocol and contrast agent injection scheme is provided in \u003cstrong\u003eSupplementary Table 1\u003c/strong\u003e. Automatic contrast agent tracking was employed with a threshold of 150 HU. The arterial phase scan was initiated with an 8-second delay, followed by the venous phase scan 35 seconds after the arterial phase. Following the scan, conventional 120 kVp images and SBI were reconstructed with a 1 mm slice thickness and 1 mm increment, including virtual monoenergetic images (VMI) 40keV, 65keV, and 100keV, iodine density (ID) maps, effective atomic number (Zeff) maps, and electron density (ED) maps.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical data collection and image analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eClinical information, such as age, gender, and smoking history, was collected for all patients. Tumor conventional imaging features were independently reviewed by two radiologists (3 and 8 years of chest imaging diagnosis experience) while blinded to pathology. Disagreements were resolved through discussion to reach a consensus. Evaluated conventional\u0026nbsp;radiological features included long-axis diameter (LD), consolidation/tumor ratio (CTR),\u0026nbsp;shape, nodule type (pGGN, mixed solid, solid nodule), lobulation, spiculation, air bronchogram, tumor-lung interface, vacuole sign, and pleural indentation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDLCT quantitative parameter measurement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOn Philips post-processing workstation (IntelliSpace Portal, Philips Healthcare), DLCT quantitative parameters were independently measured by two radiologists. Final values were calculated as the average of both measurements. The largest axial plane of the lesion was identified on conventional 120 kVp lung window images (window width 1600 HU, window level -600 HU) in the venous-phase. The region of interest (ROI) for the tumor was then manually outlined along the lesion contour. Large blood vessels, calcification, necrosis, and cavities were carefully avoided when defining the ROI. To maintain consistency across all energy images, the workstation\u0026apos;s \u0026quot;copy and paste\u0026quot; function was utilized. Measured quantitative parameters include spectral curve slope (\u0026lambda;HU), normalized iodine density (NID), Zeff, and ED value. Additionally, to calculate NID, the ID of the aorta was also recorded at the same axial plane as the tumor. \u003cstrong\u003eSupplemental Figure 1\u003c/strong\u003e presents a schematic for DLCT quantitative parameter measurement. \u0026lambda;HU and NID\u0026nbsp;were calculated as:\u003c/p\u003e\n\u003cp\u003e\u0026lambda;HU = (CT\u003csub\u003e40keV\u0026nbsp;\u003c/sub\u003e- CT\u003csub\u003e100keV\u003c/sub\u003e)/100 - 40\u003c/p\u003e\n\u003cp\u003eNID = ID\u003csub\u003etumor\u003c/sub\u003e /ID\u003csub\u003eaorta\u003c/sub\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTumor delineation and radiomics feature extraction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA radiologist manually delineated the three-dimensional ROI of the tumor on the conventional venous-phase images via ITK-SNAP (version 3.8.0). Radiomic features were extracted from seven type images: conventional, VMI at 40keV, 65keV, and 100keV, ID, Zeff, and ED. Since all DLCT energy images had the same dimensions, tumor ROI delineation was performed solely on the conventional image, which could be perfectly aligned with the reconstructed energy images. All images were standardized to a voxel size of 1 \u0026times; 1 \u0026times; 1 mm\u0026sup3;, and a bin width of 25 was implemented to ensure uniform intensity scaling. The Pyradiomics package (version 3.6.0) was used to extract radiomic features, including first-order, shape, and texture features. Detailed information of these features \u003cstrong\u003eis provided\u003c/strong\u003e in \u003cstrong\u003eSupplemental Appendix S1\u003c/strong\u003e. Additionally, another radiologist randomly selected 30 patients \u003cstrong\u003eto\u003c/strong\u003e delineate the ROI \u003cstrong\u003eand\u003c/strong\u003e assess inter-observer reproducibility by calculating the inter-class correlation coefficient (ICC).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFeature selection and model construction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRadiomic features with an ICC \u0026ge; 0.75, reflecting good reproducibility, were retained for further analysis. These features were subsequently standardized to Z-scores to ensure comparability across different scales. A three-step approach was employed to refine feature selection within the training dataset. First, features with \u003cem\u003eP\u003c/em\u003e-value \u0026lt; 0.05 were selected using the Mann-Whitney U test. Second, to address collinearity among features, Pearson\u0026apos;s correlation analysis was performed. When the correlation coefficient between two features exceeded 0.9, one of the correlated features was excluded. Finally, the least absolute shrinkage and selection operator (LASSO) algorithm, with 5-fold cross-validation, was applied to identify the optimal feature set. Based on these selected features from each type image, radiomics models were constructed using a logistic regression (LR) machine learning classifier, resulting in seven distinct radiomic signatures (Rad-sig).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo develop a clinical-radiological model, univariate and multivariate logistic regression analyses were performed. This model incorporated relevant clinical information, traditional imaging characteristics, and DLCT parameters to deliver a holistic evaluation. Finally, the Rad-sig from the best-performing radiomics model was combined with independent predictive factors from the clinical-radiological model to develop a nomogram. The entire study workflow is depicted in \u003cstrong\u003eFigure\u0026nbsp;2\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eStatistical analyses were performed using Python\u0026nbsp;(version 3.7.0),\u0026nbsp;SPSS\u0026nbsp;(version 26.0) and R (version 4.1.0)\u0026nbsp;software. Continuous variables were assessed for normality with the Shapiro-Wilk test and for variance homogeneity with the Levene test. Normally distributed variables were analyzed using parametric tests (t-test or analysis of variance), while nonparametric alternatives (Mann-Whitney U test and Kruskal-Wallis test) were applied to non-normally distributed variables.\u0026nbsp;Categorical data were analyzed using either the Chi-square or Fisher\u0026rsquo;s exact tests. For multiple comparisons, the Bonferroni correction method was used for adjustment. Statistical significance was defined as \u003cem\u003eP\u003c/em\u003e\u0026lt;0.05. Receiver operating characteristic (ROC) curve analysis was performed to evaluate model performance in predicting STAS, with calculation of sensitivity, specificity, accuracy, and area under the curve (AUC). The DeLong test was used to compare AUC values between models. Calibration curves and the Hosmer-Lemeshow test assessed agreement between predicted probabilities and observed outcomes. Decision curve analysis (DCA) evaluated the clinical net benefit of the models.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003ePatient characteristics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 266 patients diagnosed with LUAD (124 Men, 142 Women; ages 35\u0026ndash;83 years) were divided into three groups: the \u003cstrong\u003etraining set\u003c/strong\u003e (131 patients, 40 STAS-positive, 91 STAS-negative), the \u003cstrong\u003einternal validation set\u003c/strong\u003e (57 patients, 18 STAS-positive, 39 STAS-negative), and the \u003cstrong\u003eexternal validation set\u003c/strong\u003e (78 patients, 23 STAS-positive, 55 STAS-negative). Except for the lobulation sign (\u003cem\u003eP\u0026nbsp;\u003c/em\u003e= 0.032), no significant differences were observed in clinical or conventional radiological features across the three cohorts. The baseline characteristics are shown in \u003cstrong\u003eTable 1\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1.\u0026nbsp;\u003c/strong\u003eClinical characteristics and radiology parameters\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"566\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 13.5671%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.8088%;\"\u003e\u003cbr\u003e\u003cbr\u003e\n \u003cp\u003e\u003cstrong\u003eTraining set\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; (n=131)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 16.2085%;\"\u003e\u003cbr\u003e\u003cbr\u003e\n \u003cp\u003e\u003cstrong\u003eInternal Validation Set\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(n=57)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eExternal Validation Set\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(n=78)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e \u003cstrong\u003evalue\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eAge\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e62(56.0, 69.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e65(59.0, 71.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e61(66.0, 69.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.245\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eGender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.198\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eMan\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e68(51.91)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e22(38.60)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e34(43.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Woman\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e63(48.09)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e35(61.40)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e44(56.41)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eSmoking history\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.323\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;No\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e85(64.89)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e37(64.91)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e58(74.36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;YES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e46(35.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e20(35.09)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e20(25.64)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eLocation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.671\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; Left upper lobe\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e29(22.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e21(36.84)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e25(32.05)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Left lower lobe\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e20(15.27)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e8(14.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e12(15.38)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Right upper lobe\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e41(31.30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e16(28.07)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e21(26.92)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Right middle lobe\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e14(10.69)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e4(7.02)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e6(7.69)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Right lower lobe\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e27(20.61)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e8(14.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e14(17.95)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eLD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.533\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026le; 20 mm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e59(45.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e30(52.63)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e40(51.28)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026gt; 20 mm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e72(54.96)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e27(47.37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e38(48.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eCTR\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e0.64(0.26, 1.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e0.63(0.34, 1.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e0.71(0.29, 1.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.812\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eNodule type\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.965\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;pGGN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e27(20.61)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e10(17.54)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e13(16.67)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eMixed solid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e57(43.51)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e26(45.62)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e36(46.15)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eSolid nodule\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e47(35.88)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e21(36.84)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e29(37.18)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eLobulation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.032\u003csup\u003eb\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Absent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e45(34.35)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e14(24.56)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e14(17.95)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Present\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e86(65.65)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e43(75.44)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e64(82.05)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eSpiculation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.071\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Absent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e85(64.89)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e33(57.89)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e38(48.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Present\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e46(35.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e24(42.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e40(51.28)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eShape\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.333\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; Round to oval\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e58(44.27)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e31(54.39)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e33(42.31)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Irregular\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e73(55.73)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e26(45.61)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e45(57.69)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eAir bronchogram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.472\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Absent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e76(58.02)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e32(56.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e51(65.38)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Present\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e55(41.98)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e25(43.86)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e27(34.62)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eTumor-lung interface\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.785\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Well-defined\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e102(77.86)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e45(78.95)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e58(74.36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Ill-defined\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e29(22.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e12(21.05)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e20(25.64)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eVacuole sign\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.948\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Absent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e103(78.63)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e45(78.95)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e60(76.92)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Present\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e28(21.37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e12(21.05)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e18(23.08)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003ePleural\u0026nbsp;indentation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.220\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Absent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e68(51.91)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e22(38.60)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e35(44.87)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;Present\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e63(48.09)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e35(61.40)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e43(55.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003e\u0026lambda;HU\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e2.06(1.71,2.48)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e2.00(1.59,2.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e2.26(1.81,2.77)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.135\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eNID\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e0.36\u0026plusmn;0.08\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e0.35\u0026plusmn;0.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e0.38\u0026plusmn;0.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.089\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eZeff\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e8.52\u0026plusmn;0.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e8.48\u0026plusmn;0.37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e8.54\u0026plusmn;0.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.616\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 17.5292%;\"\u003e\n \u003cp\u003eED\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.8088%;\"\u003e\n \u003cp\u003e89.50(67.85, 98.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 16.2085%;\"\u003e\n \u003cp\u003e88.00(72.10, 98.30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 12.0063%;\"\u003e\n \u003cp\u003e92.40(71.40, 97.80)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 5.2828%;\"\u003e\n \u003cp\u003e0.950\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003csup\u003ea\u0026nbsp;\u003c/sup\u003eData were presented as medians (25th-75th percentiles) and compared using the Kruskal-Wallis rank sum test.\u003c/p\u003e\n\u003cp\u003e\u003csup\u003eb\u003c/sup\u003e Statistical significance was found between the training and external validation sets.\u003c/p\u003e\n\u003cp\u003eLD long-axis diameter, CTR consolidation/tumor ratio, pGGN pure ground-glass nodule, \u0026lambda;HU spectral curve slope, NID normalized iodine density, Zeff effective atomic number, ED electron density.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical-radiological model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUnivariate logistic regression analysis of the training dataset revealed several clinical and radiological factors associated with STAS (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05). These factors included tumor LD, CTR, nodule type, tumor-lung interface, pleural indentation, Zeff value, and ED value (\u003cstrong\u003eTable 2\u003c/strong\u003e). Following multivariate analysis, tumor-lung interface (OR: 3.54; 95% CI: 1.09\u0026ndash;11.46; \u003cem\u003eP\u0026nbsp;\u003c/em\u003e= 0.035) and ED value (OR: 1.17; 95% CI: 1.02\u0026ndash;1.34; \u003cem\u003eP\u0026nbsp;\u003c/em\u003e= 0.028) were identified as independent predictors of STAS. The clinical-radiological model demonstrated satisfactory predictive performance, with AUCs of 0.870 (95% CI: 0.805\u0026ndash;0.935) in the training set, 0.798 (95% CI: 0.675\u0026ndash;0.922) in the internal validation set, and 0.819 (95% CI: 0.717\u0026ndash;0.921) in the external validation set.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2.\u0026nbsp;\u003c/strong\u003eUnivariate and multivariate logistic regression analyses of\u0026nbsp;clinical-radiological\u0026nbsp;characteristics in training set\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"99%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariables\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 220px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eUnivariate analysis\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 194px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMultivariate analysis\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 70px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 70px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOR (95%CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003evalue\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOR (95%CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003evalue\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.02 (0.98 ~ 1.06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.363\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eGender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.72 (0.81 ~ 3.66)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.155\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eSmoking history\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.58 (0.73 ~ 3.40)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.242\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eLocation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp; Left upper lobe\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.00 (Reference)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp; Left lower lobe\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e2.10 (0.61 ~ 7.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp; Right upper lobe\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.46 (0.50 ~ 4.28)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.491\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp; Right middle lobe\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e2.36 (0.61 ~ 9.16)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.216\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026nbsp; Right lower lobe\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e0.90 (0.26 ~ 3.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.865\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eLD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e3.50 (1.53 ~ 7.99)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.003\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 134px;\"\u003e\n \u003cp\u003e1.54 (0.53 ~ 4.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e0.431\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 139px;\"\u003e\n \u003cp\u003eCTR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e142.77(18.92 ~ 1077.15)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt;0.001\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 134px;\"\u003e\n \u003cp\u003e1.93 (0.02 ~ 242.39)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e0.791\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eNodule type\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003epGGN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.00 (Reference)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e1.00 (Reference)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eMixed solid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e5.53 (0.67 ~ 45.62)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.112\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 134px;\"\u003e\n \u003cp\u003e0.02 (0.00 ~ 1.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e0.072\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eSolid nodule\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e41.89 (5.23 ~ 335.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt;0.001\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 134px;\"\u003e\n \u003cp\u003e0.01 (0.00 ~ 4.01)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e0.139\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eLobulation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.87 (0.82 ~ 4.30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.138\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eSpiculation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.16 (0.54 ~ 2.52)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.705\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eShape\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e2.03 (0.93 ~ 4.43)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.074\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eAir bronchogram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e0.56 (0.26 ~ 1.22)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.147\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" style=\"width: 139px;\"\u003e\n \u003cp\u003eTumor-lung interface\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 159px;\"\u003e\n \u003cp\u003e3.30 (1.40 ~ 7.77)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 61px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.006\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 134px;\"\u003e\n \u003cp\u003e3.54 (1.09 ~ 11.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.035\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eVacuole sign\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.65 (0.69 ~ 3.95)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.259\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003ePleural\u0026nbsp;indentation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e2.71 (1.25 ~ 5.87)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.011\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 134px;\"\u003e\n \u003cp\u003e0.74 (0.25 ~ 2.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e0.594\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003e\u0026lambda;HU\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e0.66 (0.35 ~ 1.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.203\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eNID\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e6.48 (0.07 ~ 591.82)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.417\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 134px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eZeff\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e0.05 (0.01 ~ 0.21)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt;0.001\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 134px;\"\u003e\n \u003cp\u003e0.61 (0.09 ~ 4.24)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e0.620\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"bottom\" style=\"width: 139px;\"\u003e\n \u003cp\u003eED\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 159px;\"\u003e\n \u003cp\u003e1.13 (1.07 ~ 1.19)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt;0.001\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 134px;\"\u003e\n \u003cp\u003e1.17 (1.02 ~ 1.34)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 60px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.028\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eLD\u0026nbsp;long-axis diameter, CTR consolidation/tumor ratio,\u0026nbsp;pGGN pure ground-glass nodule,\u0026nbsp;\u0026lambda;HU\u0026nbsp;spectral curve slope,\u0026nbsp;NID\u0026nbsp;normalized iodine density,\u0026nbsp;Zeff\u0026nbsp;effective atomic number,\u0026nbsp;ED\u0026nbsp;electron density.\u003c/p\u003e\n\u003cp\u003e* p \u0026lt; 0.05\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFeature selection and performance of radiomics model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 1,834 radiomic features were extracted from seven imaging modalities: Conventional, VMI 40keV, 65keV, 100keV, ID, Zeff, and ED. LASSO regression identified\u0026nbsp;8-19 features with non-zero coefficients for multi-parametric DLCT imaging, and their feature-weight distributions are depicted in \u003cstrong\u003eSupplemental Figure 2\u003c/strong\u003e. An LR classifier was then employed to construct the radiomics model. \u003cstrong\u003eFigure 3\u003c/strong\u003e illustrates ROC curves for the seven models, and \u003cstrong\u003eTable 3\u003c/strong\u003e summarizes their predictive metrics. The models exhibited AUC ranges of 0.847\u0026ndash;0.945 (training), 0.751\u0026ndash;0.835 (internal validation), and 0.768\u0026ndash;0.828 (external validation).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo identify the most suitable model for diagnostic development, we compared the performance of different sequences within the radiomics models. Among these, the VMI 40keV-based radiomics model demonstrated the best predictive performance, achieving an AUC of 0.899 (95% CI: 0.843\u0026ndash;0.955) in the training set, 0.835 (95% CI: 0.711\u0026ndash;0.958) in the internal validation set, and 0.828 (95% CI: 0.736\u0026ndash;0.921) in the external validation set. Given its consistently superior performance across all datasets, the VMI 40keV model was selected as the final model for nomogram construction.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 3.\u003c/strong\u003e Performance\u0026nbsp;of different radiomics models in predicting STAS status of\u0026nbsp;lung adenocarcinoma\u0026nbsp;in the training set, internal validation set, and external validation set\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" align=\"left\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAUC (95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSensitivity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSpecificity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTraining set\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 13px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eConventional\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.885 (0.830 - 0.941)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.763\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.900\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.703\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI\u0026nbsp;40keV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.899 (0.843 - 0.955)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.802\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.825\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.791\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI\u0026nbsp;65keV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.894 (0.839 - 0.948)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.832\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.850\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.824\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI 100keV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.910 (0.861 - 0.959)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.847\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.800\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.868\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u0026nbsp; Iodine density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.945 (0.911 - 0.980)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.870\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.950\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.835\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u0026nbsp; Zeff\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.847 (0.773 - 0.921)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.824\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.650\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.901\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eElectron density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.890 (0.834 - 0.947)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.786\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.900\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.736\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eInternal validation set\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eConventional\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.775 (0.646 - 0.904)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.719\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.722\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.718\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI\u0026nbsp;40keV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.835 (0.711 - 0.958)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.789\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.833\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.769\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI\u0026nbsp;65keV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.793 (0.663 - 0.924)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.702\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.833\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.641\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI 100keV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.799 (0.673 - 0.925)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.719\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.692\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u0026nbsp; Iodine density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.751 (0.610 - 0.891)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.772\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.389\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.949\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u0026nbsp; Zeff\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.825 (0.701 - 0.948)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.772\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.833\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.744\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eElectron density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.812 (0.682 - 0.942)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.754\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.667\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.795\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eExternal validation set\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eConventional\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.801 (0.704 - 0.898)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.705\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.826\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.655\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI\u0026nbsp;40keV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.828 (0.736 - 0.921)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.782\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.739\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.800\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI\u0026nbsp;65keV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.804 (0.703 - 0.904)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.718\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.783\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.691\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI 100keV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.774 (0.665 - 0.883)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.718\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.696\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.727\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u0026nbsp; Iodine density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.795 (0.694 - 0.896)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.756\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.783\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.745\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u0026nbsp; Zeff\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.769 (0.655 - 0.884)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.756\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.783\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.745\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eElectron density\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.794 (0.690 - 0.899)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.756\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.826\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.727\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 169px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 25px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 27px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 74px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 80px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 80px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eSTAS\u0026nbsp;spread through air spaces, AUC area under the curve, CI c\u003cstrong\u003eonfidence interval,\u0026nbsp;\u003c/strong\u003eVMI virtual monoenergetic images, Zeff effective atomic number.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNomogram development and validation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA nomogram (\u003cstrong\u003eFigure 4A\u003c/strong\u003e) was developed by integrating independent predictors (tumor-lung interface and ED value) with the Rad-sig derived from the optimal radiomics model (VMI 40keV). Compared with the clinical-radiological model, the nomogram achieved higher AUCs for distinguishing STAS (0.910 vs 0.870, training set; 0.868 vs 0.798, internal validation set; 0.848 vs 0.819, external validation set). In the training and internal validation sets, the nomogram showed significantly higher performance than the clinical-radiological model (\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u0026lt; 0.05), whereas no significant improvement was observed in the external validation set (\u003cem\u003eP\u0026nbsp;\u003c/em\u003e= 0.184). Furthermore, the performance of the nomogram did not differ significantly from that of the radiomics model across all three datasets (\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u0026gt; 0.05). \u003cstrong\u003eTable 4\u003c/strong\u003e summarizes the predictive performance of each model, with the associated ROC curves shown in \u003cstrong\u003eFigure 4B-D\u003c/strong\u003e. Results of the DeLong test for each model are provided in \u003cstrong\u003eSupplemental Figure 3\u003c/strong\u003e. As shown in \u003cstrong\u003eFigure 5A-C\u003c/strong\u003e, the nomogram demonstrated satisfactory calibration across the three cohorts. The Hosmer\u0026ndash;Lemeshow test indicated that the nomogram predictions (training set: \u003cem\u003eP\u0026nbsp;\u003c/em\u003e= 0.383; internal validation set: \u003cem\u003eP\u0026nbsp;\u003c/em\u003e= 0.125; external validation set: \u003cem\u003eP\u0026nbsp;\u003c/em\u003e= 0.083) closely aligned with the actual outcomes, with no significant difference. Decision curve analysis further revealed that, across all datasets, the net benefits of the clinical-radiological, radiomics, and nomogram models generally surpassed those of either treating all patients or not across most reasonable threshold ranges. Additionally, when the risk threshold was less than 0.5, the clinical-radiological model demonstrated a relatively lower overall net benefit (\u003cstrong\u003eFigure 5D-F\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 4.\u0026nbsp;\u003c/strong\u003ePredictive performance of\u0026nbsp;clinical-radiologic model, radiomics model, and nomogram\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" align=\"left\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAUC (95% CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAccuracy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSensitivity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSpecificity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTraining set\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 13px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eClinical-radiological\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.870 (0.805 - 0.935)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.779\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.900\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.725\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI\u0026nbsp;40KeV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.899 (0.843 - 0.955)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.802\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.825\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.791\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eNomogram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.910 (0.856 - 0.965)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.855\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.875\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.846\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eInternal validation set\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eClinical-radiological\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.798 (0.675 - 0.922)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.702\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.667\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI\u0026nbsp;40KeV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.835 (0.711 - 0.958)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.789\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.833\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.769\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eNomogram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.868 (0.757 - 0.978)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.842\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.778\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.872\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eExternal validation set\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eClinical-radiological\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.819 (0.717 - 0.921)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.756\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.783\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.745\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eVMI\u0026nbsp;40KeV\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.828 (0.736 - 0.921)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.782\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.739\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.800\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 29px;\"\u003e\n \u003cp\u003eNomogram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 25px;\"\u003e\n \u003cp\u003e0.848 (0.752 - 0.944)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 14px;\"\u003e\n \u003cp\u003e0.795\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.826\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15px;\"\u003e\n \u003cp\u003e0.782\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eAUC area under the curve, CI c\u003cstrong\u003eonfidence interval,\u0026nbsp;\u003c/strong\u003eVMI virtual monoenergetic images.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eImproving non-invasive and objective methods for distinguishing aggressive subtypes of LUAD, such as STAS, holds significant clinical value. This study evaluated the diagnostic performance of multi-parametric DLCT for identifying STAS. In this dual-center study, we found that ED values and tumor-lung interface were independent predictors for identifying STAS. The \u003cstrong\u003enomogram\u003c/strong\u003e that incorporated these two predictive factors, along with the \u003cstrong\u003eVMI 40keV radiomics signature\u003c/strong\u003e, outperformed the clinical-radiological model, providing\u0026nbsp;superior diagnostic accuracy\u0026nbsp;of STAS status in LUAD.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe incidence of STAS in LUAD has been found to vary widely across different studies, with reported rates ranging from 15.7% to 58.4%\u003c/strong\u003e[13,22]. \u003cstrong\u003eIn this study, the overall STAS incidence was 30.5%, consistent with previously published studies.\u003c/strong\u003ePrevious studies have shown that certain CT morphological characteristics, such as nodule type, spiculation, air bronchogram, tumor-lung interface and pleural indentation, are related to STAS status[2,13,23,24]. The results of this study are generally consistent with these findings, but there are slight differences in spiculation and air bronchogram. The analysis of quantitative parameters, revealed that STAS was significantly related to both increased tumor LD and increased CTR. The work of Kim\u0026nbsp;et al[2]\u0026nbsp;and\u0026nbsp;Tasnim et al[25]\u0026nbsp;also supports this result. However, when ED was incorporated into the multivariate analysis, the results indicated that ED was an independent predictor of STAS, whereas CTR was excluded. We speculate that ED, as a DLCT quantitative parameter, may reflect the physical density of the tumor, thereby providing insights into its pathological characteristics[26]. Tumors with a high proportion of solid components typically present higher CT values, which could explain the observed correlation between these characteristics. Therefore, we hypothesize that ED can characterize STAS more effectively than CTR. A study conducted by Liu et al[17]\u0026nbsp;supported this conclusion, as their model, which utilized LD and venous-phase ED values in 225 patients with LUAD, demonstrated satisfactory performance (AUC = 0.840). However, the absence of an independent validation set limits the assessment of its external applicability. In contrast, the clinical-radiological model in our study, which incorporates tumor-lung interface and ED values, exhibited robust applicability and reliability across diverse datasets.\u003c/p\u003e\n\u003cp\u003eMost radiomics-related studies have extracted features from conventional CT and PET/CT\u0026nbsp;images to build models[27,28]. Recently, studies have shifted their focus to the importance of DLCT images in radiomics, suggesting that DLCT image-derived features could enhance model performance[29,30]. Radiomic features were extracted from a variety of imaging modalities in this study, including conventional images, VMI at 40keV, 65keV, and 100keV, as well as ID, Zeff, and ED maps. Consequently, seven distinct radiomics models were developed. Among these, the model based on VMI 40keV demonstrated the best performance. Several studies have confirmed that, compared to conventional reconstructed images, VMI 40keV provides superior tissue contrast, thereby enhancing the accuracy of lesion detection[31,32]. A previous DLCT imaging study on thoracic tumors demonstrated that VMI in the energy range of 40 to 70keV have a significantly higher signal-to-noise ratio (SNR) compared to conventional reconstructed images, with the SNR progressively improving as the energy level decreases[33]. In terms of subjective visual assessment, VMI at 40keV outperformed both 70keV and conventional reconstructed images[33]. Wang et al found that among VMI 40-140keV (with intervals of 10keV), the deep learning model based on VMI 40keV outperformed other energy levels in predicting lymph node metastasis in lung cancer[34], which is consistent with our findings. This suggests that imaging at lower energy levels may provide more detailed information about tumor angiogenesis and heterogeneity[34]. These results provide a theoretical foundation for the superiority of the radiomics model based on VMI 40keV compared to other images.\u003c/p\u003e\n\u003cp\u003eAdditionally, we developed a nomogram that integrates conventional imaging features, DLCT quantitative parameters, and radiomics signature from VMI 40keV to predict STAS. This nomogram achieved optimal predictive performance, significantly outperforming the clinical-radiological model in both the training and internal validation sets (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05). Although the improvement was not statistically significant in the external validation set (\u003cem\u003eP\u003c/em\u003e \u0026gt; 0.05), the nomogram still achieved higher AUC, accuracy, sensitivity, and specificity than the clinical-radiological model. This enhanced performance may stem from the nomogram’s ability to reveal potential heterogeneity in LUAD across multiple dimensions[35].\u003c/p\u003e\n\u003cp\u003eThis study has certain limitations. First, the relatively small sample size of this retrospective study may introduce potential selection and statistical bias. Second, images were obtained exclusively using the Philips Spectral CT scanner, so the predictive performance and reproducibility of the model with DLCT images acquired from other manufacturers require further validation. Finally, combining radiomics models from both tumor and peritumoral regions may enhance the nomogram’s predictive performance. To validate the predictive significance of peritumoral information, large-scale, multicenter prospective studies are essential.\u003c/p\u003e\n\u003cp\u003eIn conclusion, w\u003cstrong\u003ee constructed and validated\u003c/strong\u003ea DLCT-based nomogram, incorporating conventional radiological features, DLCT quantitative parameters, and VMI 40keV radiomic features. This method represents an effective, non-invasive, and feasible auxiliary diagnostic tool that aids clinicians in preoperative prediction of STAS, supporting clinical decision-making for patients with LUAD.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eDLCT: Dual-layer spectral detector CT\u003c/p\u003e\n\u003cp\u003eSTAS: Spread through air spaces\u003c/p\u003e\n\u003cp\u003eLUAD: Lung adenocarcinoma\u003c/p\u003e\n\u003cp\u003eVMI: Virtual monoenergetic images\u003c/p\u003e\n\u003cp\u003eID: Iodine density\u003c/p\u003e\n\u003cp\u003eZeff: Effective atomic number\u003c/p\u003e\n\u003cp\u003eED: Electron density\u003c/p\u003e\n\u003cp\u003eSBI: Spectral-based images\u003c/p\u003e\n\u003cp\u003eLD: Long-axis diameter\u003c/p\u003e\n\u003cp\u003eCTR: Consolidation/tumor ratio\u003c/p\u003e\n\u003cp\u003eROI: Region of interest\u003c/p\u003e\n\u003cp\u003e\u0026lambda;HU: Spectral curve slope\u003c/p\u003e\n\u003cp\u003eICC: Inter-class correlation coefficient\u003c/p\u003e\n\u003cp\u003eLASSO: Least absolute shrinkage and selection operator\u003c/p\u003e\n\u003cp\u003eLR: Logistic regression\u003c/p\u003e\n\u003cp\u003eAUC: Area under the curve\u003c/p\u003e\n\u003cp\u003eDCA: Decision curve analysis\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe sincerely thank all the individuals who participated in these studies, as well as the researchers and technicians whose contributions made this work possible.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by the National Natural Science Foundation of China (grant numbers:82160335, 82202140) and the National Key R\u0026amp;D Program of China (grant number: 2022YFC2010002).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePei Huang, Ze Lin and Yingying Qiu contributed equally to this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors and Affiliations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDepartment of Radiology, Jiangxi Provincial People\u0026rsquo;s Hospital, The First Affiliated Hospital of Nanchang Medical College, Nanchang 330006, China\u003c/p\u003e\n\u003cp\u003ePei Huang, Yingying Qiu, Dan Chen \u0026amp; Bing Fan\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDepartment of Radiology, Wuhan Third Hospital, Tongren Hospital of Wuhan University, Wuhan, Hubei\u003c/strong\u003e, China\u003c/p\u003e\n\u003cp\u003eZe Lin \u0026amp; Yiqing Tan\u003c/p\u003e\n\u003cp\u003eDepartment of Radiology, the Second Affiliated Hospital of Naval Medical University, Shanghai 200003, China\u003c/p\u003e\n\u003cp\u003eLi Fan\u003c/p\u003e\n\u003cp\u003eDepartment of Radiology, the Affiliated Hospital of Guizhou Medical University, Guiyang 550004, China\u003c/p\u003e\n\u003cp\u003ePinggui Lei\u003c/p\u003e\n\u003cp\u003eClinical \u0026amp; Technical Support, Philips Healthcare, Shanghai, People\u0026rsquo;s Republic of China\u003c/p\u003e\n\u003cp\u003eYuting Liao\u003c/p\u003e\n\u003cp\u003eDepartment of Radiology, The Second Affiliated Hospital, Jiangxi Medical College, Nanchang University, Nanchang 330006, China\u003c/p\u003e\n\u003cp\u003eJiangxi Provincial Key Laboratory of Intelligent Medical Imaging, Nanchang 330006, China\u003c/p\u003e\n\u003cp\u003eMinjing Zuo\u003c/p\u003e\n\u003cp\u003eContributions\u003c/p\u003e\n\u003cp\u003eP.H., Z.L., Y.Q. and B.F. conceptualized the study. P.H., Z.L., D.C. and Y.T.\u0026nbsp;organized the database. P.H. and Z.L. wrote the manuscript. P.H., Z.L. and Y.L. performed the statistical analyses. L.F., P.L. and M.Z. provided critical feedback and discussions. B.F. edited the manuscript. M.Z. and B.F. supervised this study. All authors have read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorresponding authors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCorrespondence to Minjing Zuo or Bing Fan.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was approved by the Ethics Committee of Jiangxi Provincial People\u0026rsquo;s Hospital (ethics approval number: KK-2024-029) and the Second Affiliated Hospital of Nanchang University (ethics approval number: IIT-O-2024-066), following the Declaration of Helsinki. Due to the retrospective nature of the research, the Ethics Committee waived the requirement for informed consent.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting\u003c/strong\u003e\u003cstrong\u003einterest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYuting Liao is employee of Philips Healthcare (Shanghai, China). The remaining authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eTravis WD, Brambilla E, Nicholson AG, Yatabe Y, Austin JHM, Beasley MB, et al. The 2015 World Health Organization Classification of Lung Tumors: Impact of Genetic, Clinical and Radiologic Advances Since the 2004 Classification. Journal of Thoracic Oncology. 2015; 10:1243-1260.\u003c/li\u003e\n\u003cli\u003eKim SK, Kim TJ, Chung MJ, Kim TS, Lee KS, Zo JI, et al. Lung Adenocarcinoma: CT Features Associated with Spread through Air Spaces. Radiology. 2018; 289:831-840.\u003c/li\u003e\n\u003cli\u003eToyokawa G, Yamada Y, Tagawa T, Oda Y. Significance of spread through air spaces in early-stage lung adenocarcinomas undergoing limited resection. Thoracic Cancer. 2018; 9:1255-1261.\u003c/li\u003e\n\u003cli\u003eBlaauwgeers H, Flieder D, Warth A, Harms A, Monkhorst K, Witte B, et al. A Prospective Study of Loose Tissue Fragments in Non-Small Cell Lung Cancer Resection Specimens: An Alternative View to \u0026quot;Spread Through Air Spaces\u0026quot;. American Journal of Surgical Pathology. 2017; 41:1226-1230.\u003c/li\u003e\n\u003cli\u003eWarth A, Muley T, Kossakowski CA, Goeppert B, Schirmacher P, Dienemann H, et al. Prognostic Impact of Intra-alveolar Tumor Spread in Pulmonary Adenocarcinoma. American Journal of Surgical Pathology. 2015; 39:793-801.\u003c/li\u003e\n\u003cli\u003eAly RG, Rekhtman N, Li X, Takahashi Y, Eguchi T, Tan KS, et al. Spread Through Air Spaces (STAS) Is Prognostic in Atypical Carcinoid, Large Cell Neuroendocrine Carcinoma, and Small Cell Carcinoma of the Lung. Journal of Thoracic Oncology. 2019; 14:1583-1593.\u003c/li\u003e\n\u003cli\u003eLaville D, Desage AL, Fournel P, Bayle-Bleuez S, Neifer C, Picot T, et al. Spread Through Air Spaces in Stage I to III Resected Lung Adenocarcinomas: Should the Presence of Spread Through Air Spaces Lead to an Upstaging? American Journal of Surgical Pathology. 2024; 48:596-604.\u003c/li\u003e\n\u003cli\u003eEguchi T, Kameda K, Lu S, Bott MJ, Tan KS, Montecalvo J, et al. Lobectomy Is Associated with Better Outcomes than Sublobar Resection in Spread through Air Spaces (STAS)-Positive T1 Lung Adenocarcinoma: A Propensity Score-Matched Analysis. Journal of Thoracic Oncology. 2019; 14:87-98.\u003c/li\u003e\n\u003cli\u003eAltorki N, Wang X, Kozono D, Watt C, Landrenau R, Wigle D, et al. Lobar or Sublobar Resection for Peripheral Stage IA Non-Small-Cell Lung Cancer. N Engl J Med. 2023; 388:489-498.\u003c/li\u003e\n\u003cli\u003eLoganadane G, Martinetti F, Mercier O, Krhili S, Riet FG, Mbagui R, et al. Stereotactic ablative radiotherapy for early stage non-small cell lung cancer: A critical literature review of predictive factors of relapse. Cancer Treatment Reviews. 2016; 50:240-246.\u003c/li\u003e\n\u003cli\u003eVillalba JA, Shih AR, Sayo T, Kunitoki K, Hung YP, Ly A, et al. Accuracy and Reproducibility of Intraoperative Assessment on Tumor Spread Through Air Spaces in Stage 1 Lung Adenocarcinomas. Journal of Thoracic Oncology. 2021; 16:619-629.\u003c/li\u003e\n\u003cli\u003eZhou F, Villalba JA, Sayo T, Narula N, Pass H, Mino-Kenudson M, et al. Assessment of the feasibility of frozen sections for the detection of spread through air spaces (STAS) in pulmonary adenocarcinoma. Mod Pathol. 2022; 35:210-217.\u003c/li\u003e\n\u003cli\u003eToyokawa G, Yamada Y, Tagawa T, Kamitani T, Yamasaki Y, Shimokawa M, et al. Computed tomography features of resected lung adenocarcinomas with spread through air spaces. J Thorac Cardiovasc Surg. 2018; 156:1670-1676.\u003c/li\u003e\n\u003cli\u003eMoore J, Remy J, Altschul E, Chusid J, Flohr T, Raoof S, et al. Thoracic Applications of Spectral CT Scan. Chest. 2024; 165:417-430.\u003c/li\u003e\n\u003cli\u003eAgostini A, Borgheresi A, Mari A, Floridi C, Bruno F, Carotti M, et al. Dual-energy CT: theoretical principles and clinical applications. Radiologia Medica. 2019; 124:1281-1295.\u003c/li\u003e\n\u003cli\u003eHa T, Kim W, Cha J, Lee YH, Seo HS, Park SY, et al. Differentiating pulmonary metastasis from benign lung nodules in thyroid cancer patients using dual-energy CT parameters. European Radiology. 2022; 32:1902-1911.\u003c/li\u003e\n\u003cli\u003eLiu BC, Ma HY, Huang J, Luo YW, Zhang WB, Deng WW, et al. Does dual-layer spectral detector CT provide added value in predicting spread through air spaces in lung adenocarcinoma? A preliminary study. European Radiology. 2024; 34:4176-4186.\u003c/li\u003e\n\u003cli\u003eWu F, Zhou H, Li F, Wang JT, Ai T. Spectral CT Imaging of Lung Cancer: Quantitative Analysis of Spectral Parameters and Their Correlation with Tumor Characteristics. Academic Radiology. 2018; 25:1398-1404.\u003c/li\u003e\n\u003cli\u003eMayerhoefer ME, Materka A, Langs G, Haggstrom I, Szczypinski P, Gibbs P, et al. Introduction to Radiomics. Journal of Nuclear Medicine. 2020; 61:488-495.\u003c/li\u003e\n\u003cli\u003ePerez-Johnston R, Araujo-Filho JA, Connolly JG, Caso R, Whiting K, Tan KS, et al. CT-based Radiogenomic Analysis of Clinical Stage I Lung Adenocarcinoma with Histopathologic Features and Oncologic Outcomes. Radiology. 2022; 303:664-672.\u003c/li\u003e\n\u003cli\u003eChoi W, Oh JH, Riyahi S, Liu CJ, Jiang F, Chen W, et al. Radiomics analysis of pulmonary nodules in low-dose CT for early detection of lung cancer. Medical Physics. 2018; 45:1537-1549.\u003c/li\u003e\n\u003cli\u003eShiono S, Endo M, Suzuki K, Yanagawa N. Spread through air spaces affects survival and recurrence of patients with clinical stage IA non-small cell lung cancer after wedge resection. Journal of Thoracic Disease. 2020; 12:2247-2260.\u003c/li\u003e\n\u003cli\u003eQi L, Xue K, Cai Y, Lu J, Li X, Li M. Predictors of CT Morphologic Features to Identify Spread Through Air Spaces Preoperatively in Small-Sized Lung Adenocarcinoma. Frontiers in Oncology. 2020; 10:548430.\u003c/li\u003e\n\u003cli\u003eZhang X, Qiao W, Shen J, Jiang Q, Pan C, Wang Y, et al. Clinical, pathological, and computed tomography morphological features of lung cancer with spread through air spaces. Transl Lung Cancer Res. 2024; 13:2802-2812.\u003c/li\u003e\n\u003cli\u003eTasnim S, Raja S, Mukhopadhyay S, Blackstone EH, Toth AJ, Barron JO, et al. Preoperative predictors of spread through air spaces in lung cancer. J Thorac Cardiovasc Surg. 2024; 168:660-669.\u003c/li\u003e\n\u003cli\u003eTatsugami F, Higaki T, Kiguchi M, Tsushima S, Taniguchi A, Kaichi Y, et al. Measurement of electron density and effective atomic number by dual-energy scan using a 320-detector computed tomography scanner with raw data-based analysis: a phantom study. J Comput Assist Tomogr. 2014; 38:824-827.\u003c/li\u003e\n\u003cli\u003eGuo R, Yan W, Wang F, Su H, Meng X, Xie Q, et al. The utility of (18)F-FDG PET/CT for predicting the pathological response and prognosis to neoadjuvant immunochemotherapy in resectable non-small-cell lung cancer. Cancer Imaging. 2024; 24:120.\u003c/li\u003e\n\u003cli\u003eDong Y, Jiang Z, Li C, Dong S, Zhang S, Lv Y, et al. Development and validation of novel radiomics-based nomograms for the prediction of EGFR mutations and Ki-67 proliferation index in non-small cell lung cancer. Quant Imaging Med Surg. 2022; 12:2658-2671.\u003c/li\u003e\n\u003cli\u003eLi M, Fan Y, You H, Li C, Luo M, Zhou J, et al. Dual-Energy CT Deep Learning Radiomics to Predict Macrotrabecular-Massive Hepatocellular Carcinoma. Radiology. 2023; 308:e230255.\u003c/li\u003e\n\u003cli\u003eXu H, Zhu N, Yue Y, Guo Y, Wen Q, Gao L, et al. Spectral CT-based radiomics signature for distinguishing malignant pulmonary nodules from benign. Bmc Cancer. 2023; 23:91.\u003c/li\u003e\n\u003cli\u003eNagayama Y, Tanoue S, Inoue T, Oda S, Nakaura T, Utsunomiya D, et al. Dual-layer spectral CT improves image quality of multiphasic pancreas CT in patients with pancreatic ductal adenocarcinoma. European Radiology. 2020; 30:394-403.\u003c/li\u003e\n\u003cli\u003eWang X, Liu D, Jiang S, Zeng X, Li L, Yu T, et al. Subjective and Objective Assessment of Monoenergetic and Polyenergetic Images Acquired by Dual-Energy CT in Breast Cancer. Korean Journal of Radiology. 2021; 22:502-512.\u003c/li\u003e\n\u003cli\u003eDoerner J, Hauger M, Hickethier T, Byrtus J, Wybranski C, Grosse HN, et al. Image quality evaluation of dual-layer spectral detector CT of the chest and comparison with conventional CT imaging. European Journal of Radiology. 2017; 93:52-58.\u003c/li\u003e\n\u003cli\u003eWang YW, Chen CJ, Wang TC, Huang HC, Chen HM, Shih JY, et al. Multi-energy level fusion for nodal metastasis classification of primary lung tumor on dual energy CT using deep learning. Computers in Biology and Medicine. 2022; 141:105185.\u003c/li\u003e\n\u003cli\u003eZhu M, Yang Z, Wang M, Zhao W, Zhu Q, Shi W, et al. A computerized tomography-based radiomic model for assessing the invasiveness of lung adenocarcinoma manifesting as ground-glass opacity nodules. Respir Res. 2022; 23:96.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-cancer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bcan","sideBox":"Learn more about [BMC Cancer](http://bmccancer.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bcan/default.aspx","title":"BMC Cancer","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Dual-layer spectral computed tomography, Spread through air spaces, Lung adenocarcinoma, Radiomics, Nomogram","lastPublishedDoi":"10.21203/rs.3.rs-7391071/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7391071/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eObjectives\u003c/h2\u003e\u003cp\u003eTo evaluate the value of quantitative parameters and radiomic features based on dual-layer spectral detector CT (DLCT) in predicting spread through air spaces (STAS) of lung adenocarcinoma (LUAD).\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eThis study analyzed 266 patients with pathologically confirmed LUAD from two medical centers. Patients from center 1 were divided into training (n\u0026thinsp;=\u0026thinsp;131) and internal validation (n\u0026thinsp;=\u0026thinsp;57) sets, while center 2 (n\u0026thinsp;=\u0026thinsp;78) formed the external validation set. Clinical data, conventional imaging features, and DLCT quantitative parameters were analyzed to develop a clinical-radiological model. Radiomic features were extracted from venous-phase images, including conventional images, virtual monoenergetic images (VMI) at 40keV, 65keV, and 100keV, along with iodine density maps, effective atomic number (Zeff) maps, and electron density (ED) maps. The best-performing radiomics model was combined with clinical-radiological predictors to create a nomogram. Model performance was evaluated through ROC analysis, calibration curves, and decision curve analysis.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eMultivariate analysis revealed that tumor-lung interface and ED values were independent predictive factors in the clinical-radiological model. The optimal radiomics model was constructed based on VMI 40keV, demonstrating AUCs of 0.899, 0.835, and 0.828 in the training, internal validation, and external validation sets, respectively. The nomogram, which incorporated the VMI 40keV radiomics signature along with tumor-lung interface and ED values, outperformed the clinical-radiological model in the training set (AUC\u0026thinsp;=\u0026thinsp;0.910 vs 0.870; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.018) and the internal validation set (AUC\u0026thinsp;=\u0026thinsp;0.868 vs 0.798; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.046). While the improvement in the external validation set was not statistically significant (AUC\u0026thinsp;=\u0026thinsp;0.848 vs 0.819; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.184).\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eThe nomogram, which integrates conventional imaging features, DLCT quantitative parameters and VMI 40keV radiomic features, serves as a valuable non-invasive tool for the preoperative assessment of STAS in LUAD.\u003c/p\u003e","manuscriptTitle":"Dual-layer spectral detector CT quantitative parameters and radiomics for predicting spread through air spaces of lung adenocarcinoma: A dual-center study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-17 08:41:29","doi":"10.21203/rs.3.rs-7391071/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-10-03T06:52:00+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-01T10:51:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"185000728922921838834408347985859457644","date":"2025-09-26T17:19:55+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-22T03:04:26+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"67014227810743456033944700563640045648","date":"2025-09-22T01:20:10+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-09-09T13:02:58+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-18T07:54:43+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-18T07:51:59+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Cancer","date":"2025-08-17T07:39:46+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-cancer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bcan","sideBox":"Learn more about [BMC Cancer](http://bmccancer.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bcan/default.aspx","title":"BMC Cancer","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d9f8ab36-4008-4fdc-b16f-dbd32be58ce0","owner":[],"postedDate":"September 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-29T16:04:05+00:00","versionOfRecord":{"articleIdentity":"rs-7391071","link":"https://doi.org/10.1186/s12885-025-15436-7","journal":{"identity":"bmc-cancer","isVorOnly":false,"title":"BMC Cancer"},"publishedOn":"2025-12-22 15:57:31","publishedOnDateReadable":"December 22nd, 2025"},"versionCreatedAt":"2025-09-17 08:41:29","video":"","vorDoi":"10.1186/s12885-025-15436-7","vorDoiUrl":"https://doi.org/10.1186/s12885-025-15436-7","workflowStages":[]},"version":"v1","identity":"rs-7391071","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7391071","identity":"rs-7391071","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}