The potential value of CT-based whole lung radiomics nomogram for predicting osteoporosis risk in COPD patients: a two-center study

The potential value of CT-based whole lung radiomics nomogram for predicting osteoporosis risk in COPD patients: a two-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 Article The potential value of CT-based whole lung radiomics nomogram for predicting osteoporosis risk in COPD patients: a two-center study Hupo Bian, Shaoqi Zhu, Huiying Qian, Jingnan Xue, Luying Qi, Xiuhua Peng, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6456138/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract To evaluate the value of CT whole-lung imaging histograms in diagnosing osteoporosis (OP) risk in patients with chronic obstructive pulmonary disease (COPD).258 COPD patients were divided into a training cohort (n = 149), an internal validation cohort (n = 64), and an external validation cohort (n = 45). Clinical data and CT results were analyzed. Imaging histologic features of the whole lung were extracted from chest CT images. Machine learning algorithms were utilized to construct the radiomics model. Multifactor logistic regression analysis was used to build the radiomics nomogram by combining independent clinical factors. ROC curves were used to analyze the predictive performance of the models.We developed a model to predict osteoporosis risk in patients with COPD by integrating imaging histologic features, as well as independent clinical risk factors. On the training set, the joint model (area under the curve [AUC], 0.811), the clinical model (AUC, 0.691), and the imaging model (AUC, 0.762). On the internal validation set, the joint model (AUC, 0.806), the clinical model (AUC, 0.724), and the imaging model (AUC, 0.765). On the external validation set, the joint model (AUC, 0.728), the clinical model (AUC, 0.656), and the imaging model (AUC, 0.718). Decision curve analysis showed that the joint model was superior to the single radiomics model with clinical factors. CT-based whole-lung radiomics nomograms are valuable in diagnosing the risk of osteoporosis in patients with COPD. Health sciences/Endocrinology/Endocrine system and metabolic diseases Health sciences/Biomarkers Health sciences/Endocrinology Health sciences/Medical research Health sciences/Risk factors chronic obstructive pulmonary disease osteoporosis radiomics predictive modeling computed tomography Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Chronic Obstructive Pulmonary Disease (COPD) is one of the leading causes of death worldwide [ 1 ], characterized by chronic airflow limitation, and in addition to its respiratory effects, COPD is often comorbid with other chronic diseases, including cardiovascular disease, osteoporosis, diabetes, lung cancer, cachexia, anemia, anxiety and depression [ 2 – 5 ]. Among these comorbidities, osteoporosis (OP) has emerged as an important comorbidity of COPD, increasing the risk of fracture, increasing patient morbidity and mortality [ 6 , 7 ], and causing a significant social burden. According to the literature, patients with COPD are more likely to develop osteoporosis than those without COPD [ 8 ], and possible factors contributing to osteoporosis in patients with COPD include systemic inflammation, use of corticosteroids, vitamin D deficiency, smoking, hypoxia, and anemia [ 9 ]. However, the link between COPD and osteoporosis is not fully understood. Bone mineral density (BMD) is widely recognized as the gold standard for the diagnosis of osteoporosis [ 10 ]. Still, clinicians do not routinely request BMD assessment in patients with COPD, and in the absence of BMD screening, CT scans are nevertheless widely used in patients with COPD. Although CT scans are not commonly used for the diagnosis of osteoporosis, opportunistic CT testing may improve the overall screening rate for osteoporosis [ 11 ] and may reduce the incidence of future fractures; therefore, it is critical to determine how to utilize CT testing in COPD patients to assess their risk of developing osteoporosis. radiomics is an emerging high-throughput method for extracting quantitative image features [ 12 , 13 ] and has been successfully used in various aspects of early diagnosis [ 14 ], staging [ 15 ], and prediction of cardiovascular disease risk [ 16 ] in COPD patients. However, no investigator has yet utilized whole-lung radiomics to identify osteoporosis in COPD patients. Therefore, to meet clinical needs, it is necessary to investigate quantitative data obtained from lung parenchyma to identify the risk of osteoporosis in COPD patients. Based on this hypothesis, we aimed to evaluate the value of CT-based whole-lung radiomics features in determining the risk of osteoporosis in patients with COPD. Materials and Methods Patient characteristics Patients and clinical data This retrospective study was approved by the ethics committees of both hospitals (No. 2025KYLL036-01) and waived the requirement for written informed consent. The study included 258 patients with COPD diagnosed by pulmonary function tests (PFT) from January 2020 to October 2024 at both centers. Inclusion criteria were as follows:(1) COPD diagnosed by PFT; (2) PFT and chest CT completed within 2 weeks; and (3) complete thin-layer (1-mm) chest CT images. Exclusion criteria were as follows:(1) combination of other chest diseases (e.g., pneumonia, pulmonary atelectasis, pulmonary nodules > 6 mm or masses and pleural effusion; (2) combination of malignant tumors; and (3) spinal implants or significant imaging changes. A total of 258 patients were included for these etiologies. To make the study more rigorous and to prevent model overfitting, 213 patients from Hospital 1 were randomly assigned to either the training cohort (n = 149) or the internal validation cohort (N = 64) in a 7:3 ratio, and 45 patients from Hospital 2 were assigned to the external validation cohort (n = 45). The training cohort was used to develop the model, while the internal and external validation cohorts were used to assess the validity of the model in clinical practice and to enhance the robustness of the study. Clinical information included age, weight, height, BMI, gender, GOLD class, smoking status, and laboratory tests such as C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), procalcitonin (PCT), albumin, alkaline phosphatase, triglycerides, platelet distribution width (PDW), arterial partial pressure of oxygen (Pa02), arterial partial pressure of carbon dioxide ( PACO2), white blood cell count, percent neutrophils, percent lymphocytes, absolute eosinophil count, erythrocyte pressure volume, erythrocyte distribution width, mean platelet volume, hemoglobin, and globulin. Laboratory factors with missing values greater than 30% were excluded due to their retrospective nature. Finally, 14 laboratory factors: C-reactive protein, albumin, Globulin, triglyceride, Alkaline Phosphatase, White blood cell count, Neutrophilic granulocyte Percentage, Percentage of lymphocytes, Absolute eosinophil count, Plateletcrit, Red blood cell distribution width, Mean platelet volume, Platelet Distribution Width, Hemoglobin were included in further analysis. Osteoporotic events were obtained by searching for GMD examinations. The presence or absence of osteoporotic events was diagnosed at the time of admission, and the interval between admission and chest CT scan was less than 1 month.The gold standard for the precise diagnosis of osteoporosis is the dual-energy X-ray absorptiometry (DXA) technique, in which a T-score of -2.5 or lower indicates the presence of osteoporosis[ 17 ]. CT image acquisition and lung function testing Participants underwent non-contrast CT scanning with Aquilion ONE TSX-301C, Somatom Force, Brillince CT 16, et al. Axial CT images of the entire chest were acquired under full inspiration (The scanning parameters are shown in Supplementary Material 1). Pulmonary function tests were performed using a CHEST multifunction spirometer HI-801 (Japan). Forced expiratory volume in 1 s (FEV1), percent predicted of forced expiratory flow in 1 s (FEV1%), and FEV1/FVC are the diagnostic criteria for COPD. The criterion is FEV1/FVC < 0.7, and FEV1 increases less than 200 mL after the use of a bronchodilator [ 18 ]. Subjects were grouped according to the global initiative for chronic obstructive lung disease [ 19 ]: GOLD I grade, FEV1/FVC < 0.7, FEV1 ≥ 80% predicted; GOLD II grade, FEV1/FVC < 0.7, 50% predicted < FEV1 ≤ 80% predicted; GOLD III grade, FEV1/FVC < 0.7, 30% predicted < FEV1 ≤ 50% predicted; GOLD IV grade, FEVI/FVC < 0.7, FEV1 < 30% predicted. Fully automated region of interest segmentation We automatically segmented the right and left lungs by the Full Lung Segmentation component on the OnekeyAI platform ( https://github.com/OnekeyAI-Platform/onekey ), and the extracted left and right lungs were merged into a region of interest (ROI) (the specific algorithm process can be found in Supplementary Material 2). The segmentation results were then independently assessed by two chest radiologists with more than 10 years of experience using ITK-SNAP software. [ 20 ] (version 3.8.0, www.itksnap.org ) to correct erroneous segmentations. Whole lung radiomics feature extraction Radiomic feature extraction was conducted using the PyRadiomics open-source Python package (version 3.7.12; https://pyradiomics.readthedocs.io ) on the OnekeyAI platform, with three features obtained: first-order features, shape, and texture feature. Radiation features were extracted using Z-score normalization. All available features implemented in PyRadiomics were extracted from the original images and filtered images, including wavelet and Laplacian of Gaussian transformations. Before feature extraction, the image underwent a three-step preprocessing operation, voxel resampling, gray discretization, and image intensity normalization, resampling the image to 1 mm *1 mm * 1 mm, and adjusting the gray value of the image to 25 gray levels, which is helpful to reduce the influence of different layer thicknesses and reduce the interference of noise. Radiomics feature selection and model construction We normalized all extracted features using Z-score normalization and performed a t-test to assess their statistical significance, retaining only those with a p-value below 0.05. To mitigate collinearity, we examined the correlations between features using Pearson's correlation coefficient, excluding any feature from pairs that exhibited a correlation coefficient above 0.9. We further refined the feature set through Lasso regression within a 10-fold cross-validation framework to identify the optimal regularization parameter, λ, effectively reducing the feature set to those most predictive and informative. Radiomics Signature: By employing LASSO for stringent feature selection, we constructed the radiomics risk model using machine learning algorithms such as Logistic Regression, Support Vector Machine, and Random Forest. Comparative analyses were performed to gauge each model's performance, and we investigated the benefits of feature fusion by integrating multi-modal features, allowing us to explore the advantages of combining multiple imaging modalities to enhance predictive accuracy. The diagnostic efficacy of our machine learning model was rigorously evaluated in the test cohort through the construction of Receiver Operating Characteristic (ROC) curves to assess its discriminative ability,furthermore, for all models, 5-fold cross-validation was conducted. The calibration of the model was analyzed using calibration curves, complemented by the Hosmer-Lemeshow goodness-of-fit test to validate its reliability. Additionally, Decision Curve Analysis (DCA) was employed to appraise the clinical utility of our predictive models, facilitating an understanding of the potential benefits in a clinical context. Metrics The diagnostic efficacy of our machine learning model was rigorously evaluated in the test cohort through the construction of Receiver Operating Characteristic (ROC) curves to assess its discriminative ability. The calibration of the model was analyzed using calibration curves, complemented by the Hosmer-Lemeshow goodness-of-fit test to validate its reliability. Additionally, Decision Curve Analysis (DCA) was employed to appraise the clinical utility of our predictive models, facilitating an understanding of the potential benefits in a clinical context. Statistical Analysis The normality of clinical features was verified using the Shapiro-Wilk test. Continuous variables were subjected to the t-test or the Mann-Whitney U test based on their distribution characteristics. Categorical variables were examined using Chi-square (χ²) tests. The baseline characteristics across all cohorts are detailed in Table 2 . Notably, p-values exceeding 0.05 between cohorts indicated no significant differences, thus confirming an unbiased division between groups. All data analyses were executed on the OnekeyAI platform version 4.9.1 utilizing Python 3.7.12. Statistical assessments were conducted with Statsmodels version 0.13.2. Radiomics feature extraction was performed via PyRadiomics version 3.0.1. Machine learning implementations, including the Support Vector Machine (SVM), were facilitated using Scikit-learn version 1.0.2. Results The flow chart for patient selection is shown in Fig. 1 . As of October 31, 2024, we included a total of 434 patients with a diagnosis of COPD, and after screening by exclusion criteria, the final cohort (n = 258) included 80 women and 178 men, with a mean age of 74.98 years for the entire cohort, and the two hospitals included 213 and 45 patients, respectively. Among them, 72 COPD patients did not have osteoporosis, while 186 COPD patients had osteoporosis. Modeling of Clinical Factors The clinical factors of the patients in the training and test sets are shown in Table 1 . Independent risk factor identification (univariate and multivariate logistic regression analysis) is shown in Table 2 . In the univariate analysis, height, hemoglobin, C-reactive protein (CRP), alkaline phosphatase (ALP), globulin, weight, age, neutrophilic granulocyte percentage, erythrocyte sedimentation rate (ESR), albumin, arterial carbon dioxide partial pressure, BMI, and other factors. Dioxide partial pressure, BMI, Vitamin D, platelet distribution width (PDW), procalcitonin, red blood cell distribution width (RDW), mean platelet volume (MPV), white blood cell distribution width (RDW), and mean platelet volume (MPV). Volume (MPV), white blood cell count (WBC), triglyceride levels, and smoking status were all associated with the outcome, as indicated by ORs significantly different from 1 and p-values < 0.05 (or 0, indicating < 0.001). However, in the multivariate analysis, only neutrophilic granulocyte percentage (OR = 1.044, p = 0.042) and platelet crit (OR = 0.0, p = 0.044) retained significance. The other features exhibited ORs close to 1 and higher p-values, suggesting that their associations with the outcome were not independent of the different variables considered in the model. of the different variables considered in the model. Finally, neutrophil percentage and platelet count were included in the construction of the clinical factors model. Table 1 Baseline characteristics of the study population Clinical factors Training cohort(n = 149) Internal validation cohort (n = 64) External vaidation cohort(n = 45) COPD without OP(n = 50) COPD with OP(n = 99) p value COPD without OP(n = 17) COPD with OP(n = 47) p value COPD without OP(n = 5) COPD with OP(n = 40) p value Age 72.14 ± 10.95 76.20 ± 8.14 0.042 71.53 ± 7.72 77.81 ± 8.76 0.002 68.00 ± 10.46 74.53 ± 9.67 0.164 Smoke 0.66 ± 0.48 0.53 ± 0.50 0.118 0.65 ± 0.49 0.32 ± 0.47 0.02 0.40 ± 0.55 0.55 ± 0.50 0.545 Height 168.26 ± 4.96 162.36 ± 8.99 < 0.001 168.71 ± 6.66 159.23 ± 7.83 < 0.001 175.20 ± 3.56 162.93 ± 8.00 0.002 Weight 66.66 ± 10.40 57.32 ± 11.94 < 0.001 64.90 ± 9.52 54.93 ± 10.12 < 0.001 75.10 ± 14.20 54.52 ± 10.22 < 0.001 BMI 23.55 ± 3.68 21.68 ± 3.69 0.004 22.76 ± 2.69 21.62 ± 3.47 0.224 24.41 ± 4.11 20.51 ± 3.20 0.017 C-Reactive Protein 58.96 ± 44.94 60.80 ± 49.63 0.966 54.35 ± 46.89 72.30 ± 48.51 0.221 69.40 ± 55.36 103.42 ± 55.23 0.155 albumin 37.85 ± 4.79 37.80 ± 4.14 0.97 40.30 ± 4.31 37.25 ± 4.91 0.027 37.46 ± 5.23 35.54 ± 4.57 0.386 Globulin 67.12 ± 35.05 66.23 ± 32.54 0.915 70.24 ± 36.52 63.09 ± 36.96 0.495 83.40 ± 24.05 55.08 ± 33.19 0.073 triglyceride 1.32 ± 0.98 1.04 ± 0.64 0.021 1.35 ± 0.72 1.17 ± 0.70 0.294 1.82 ± 0.63 0.97 ± 0.35 0.003 Alkaline Phosphatase 86.88 ± 67.20 87.72 ± 36.65 0.24 84.29 ± 21.29 89.94 ± 42.25 0.909 77.78 ± 18.90 78.03 ± 22.23 0.899 White blood cell count 6.47 ± 2.71 6.91 ± 3.05 0.359 6.96 ± 2.88 7.57 ± 3.86 0.508 7.42 ± 1.60 7.62 ± 2.62 0.866 Neutrophilic granulocyte percentage 66.31 ± 12.64 73.67 ± 12.36 < 0.001 67.34 ± 11.81 75.39 ± 9.98 0.009 65.36 ± 4.62 70.77 ± 16.24 0.073 Percentage of lymphocytes 24.16 ± 11.65 17.60 ± 8.35 < 0.001 23.61 ± 9.44 16.46 ± 7.80 0.003 22.98 ± 4.34 18.26 ± 12.46 0.093 Absolute eosinophil count 0.19 ± 0.24 0.14 ± 0.42 < 0.001 0.13 ± 0.12 0.09 ± 0.12 0.073 0.15 ± 0.13 0.36 ± 1.24 0.814 Plateletcrit 0.20 ± 0.06 0.18 ± 0.06 0.256 0.19 ± 0.06 0.18 ± 0.07 0.35 0.19 ± 0.09 0.21 ± 0.06 0.662 Red blood cell distribution width 13.46 ± 1.18 13.49 ± 1.18 0.904 13.44 ± 0.95 13.75 ± 1.64 0.873 12.90 ± 0.81 13.65 ± 1.46 0.233 Mean platelet volume 10.10 ± 1.09 10.17 ± 1.22 0.707 10.56 ± 2.02 10.18 ± 1.24 0.808 10.58 ± 1.75 10.61 ± 1.19 0.96 Platelet Distribution Width 16.24 ± 0.96 16.28 ± 0.83 0.99 15.89 ± 1.24 16.53 ± 3.51 0.502 15.38 ± 3.02 12.37 ± 2.82 0.053 Hemoglobin 126.66 ± 14.83 121.09 ± 18.09 0.03 129.24 ± 14.63 119.36 ± 19.45 0.062 141.40 ± 22.61 123.42 ± 17.19 0.039 GOLD grade 0.251 0.393 0.132 GOLDⅠ 23(46.00) 37(37.37) 7(41.18) 14(29.79) 3(60.00) 7(17.50) GOLDⅡ 17(34.00) 27(27.27) 4(23.53) 10(21.28) 1(20.00) 8(20.00) GOLDⅢ 8(16.00) 19(19.19) 6(35.29) 18(38.30) 0 19(47.50) GOLDⅣ 2(4.00) 16(16.16) 0 5(10.64) 1(20.00) 6(15.00) Gender < 0.001 0.001 0.371 Female 4(8.00) 37(37.37) 1(5.88) 26(55.32) 0 12(30.00) Male 46(92.00) 62(62.63) 16(94.12) 21(44.68) 5(100.00) 28(70.00) Table 2 Univariable and multivariable logistic regression analysis Variable Univariable analysis Multivariable analysis OR[95%CI] p value OR[95%CI] p value Plateletcrit 19.668[4.581,84.437] 0.001 0[0,0.23] 0.044 Hemoglobin 1.005[1.003,1.007] < 0.001 0.989[0.962,1.017] 0.519 C-Reactive Protein 1.007[1.004,1.011] 0.001 0.996[0.988,1.004] 0.44 Alkaline Phosphatase 1.007[1.003,1.01] 0.001 0.997[0.989,1.005] 0.512 Globulin 1.008[1.004,1.012] 0.001 1.003[0.99,1.015] 0.714 Neutrophilic granulocyte percentage 1.011[1.007,1.015] < 0.001 1.044[1.008,1.08] 0.042 albumin 1.018[1.01,1.025] < 0.001 1.053[0.951,1.166] 0.401 Platelet Distribution Width 1.043[1.025,1.062] < 0.001 1.229[0.768,1.966] 0.471 Red blood cell distribution width 1.052[1.029,1.075] < 0.001 0.887[0.641,1.228] 0.544 Mean platelet volume 1.07[1.04,1.1] < 0.001 0.849[0.587,1.226] 0.463 White blood cell count 1.101[1.057,1.147] < 0.001 1.154[0.994,1.34] 0.115 triglyceride 1.325[1.068,1.642] 0.031 0.696[0.412,1.175] 0.255 GOLD 1.415[1.232,1.624]] < 0.001 1.178[0.856,1.624] 0.4 Smoke 1.576[1.093,2.273]] 0.041 1.116[0.48,2.591] 0.831 Height 1.004[1.002,1.006] < 0.001 1.001[0.923,1.084] 0.986 Weight 1.009[1.004,1.013] 0.002 0.879[0.771,1.003] 0.109 Age 1.01[1.006,1.014] < 0.001 1.022[0.976,1.069] 0.439 BMI 1.027[1.014,1.04] 0.001 1.264[0.884,1.806] 0.281 Feature extraction , selection, and radiomics signature construction After extracting, normalizing, and deleting features with a correlation greater than 0.9 from CT images, a total of 107 radiological features were retained. We will select features using both mRMR and LASSO. First, mRMR was performed to eliminate redundant and irrelevant features, and 15 features were retained. Then, LASSO is performed to select an optimized subset of features to construct the final model (Fig. 2 a and b). Finally, six features were used to build the radiomics feature signature (Fig. 2 c). Radiomics features were screened using the following formula. The rad score is shown as follows. Radscore = 0.6621621621621622 -0.095484 * original_shape_SurfaceArea + 0.080137 * original_glszm_SizeZoneNonUniformityNormalized + 0.016017 * original_glrlm_LongRunHighGrayLevelEmphasis + 0.004569 * original_ngtdm_Strength − 0.030314 * original_glszm_SmallAreaLowGrayLevelEmphasis − 0.074244 * original_shape_Maximum3DDiameter Development of radiological column line diagrams and evaluation of the performance of different models Table 3 displays the performance of the clinical model, radiomics signature, and nomogram in the diagnosis of training, internal, and external validation sets. Figure 3 presents that the Clinic, Radiomics, and Combined indicators are presented across the training, validation, and test cohorts. In the training cohort, the Combined indicator yields the highest AUC of 0.811, followed by Radiomics (0.762) and Clinic (0.691). In the validation cohort, the combined indicator maintains a strong performance with an AUC of 0.806, while Radiomics achieves a slightly higher AUC of 0.765 compared to Clinic (0.724)c In the test cohort, the Combined indicator demonstrates an AUC of 0.728, marginally higher than Radiomics (0.718) and significantly higher than Clinic ( 0.656). Notably, the confidence intervals for the AUCs, particularly in the test cohort, are wide, indicating variability in performance. Table 3 Metrics on different signature Signature Accuracy AUC 95% CI Sensitivity Specificity PPV NPV Cohort Clinic 0.649 0.691 0.5985–0.7827 0.602 0.74 0.819 0.487 train Radiomics 0.75 0.762 0.6814–0.8418 0.847 0.56 0.79 0.651 train Combined 0.696 0.811 0.7425–0.8795 0.612 0.86 0.896 0.531 train Clinic 0.683 0.724 0.5881–0.8594 0.63 0.824 0.906 0.452 val Radiomics 0.651 0.765 0.6427–0.8867 0.543 0.941 0.962 0.432 val Combined 0.73 0.806 0.6873–0.9239 0.696 0.824 0.914 0.5 val Clinic 0.659 0.656 0.4009–0.9119 0.641 0.8 0.962 0.222 test Radiomics 0.841 0.718 0.4065–1.0000 0.872 0.6 0.944 0.375 test Combined 0.886 0.728 0.3582–1.0000 0.923 0.6 0.947 0.5 test The analysis reveals that the Combined indicator consistently outperforms both the Clinic and Radiomics indicators in terms of AUC in the training and validation cohorts. Although the Combined indicator retains a higher AUC in the test cohort compared to the individual indicators, the wide confidence intervals suggest potential variability in its performance. Although the Combined indicator retains a higher AUC in the test cohort compared to the individual indicators, the wide confidence intervals suggest potential variability in its performance. The Radiomics indicator shows a notable increase in specificity across all cohorts, peaking at 0.941 in the validation cohort, but this is accompanied by lower sensitivity. The Clinic indicator demonstrates moderate performance but consistently exhibits lower AUC values and specificity compared to the other indicators. Overall, the Combined indicator appears to be the most robust across different cohorts, balancing the AUC values and specificity across different cohorts. Overall, the Combined indicator is the most robust across different cohorts, balancing both sensitivity and specificity, despite some variability in performance indicated by the wide confidence intervals. The Combined indicator appears to be the most robust across different cohorts, balancing both sensitivity and specificity, despite some variability in performance indicated by the wide confidence intervals. Calibration Curve The Hosmer-Lemeshow (HL) test quantifies the discrepancy between predicted probabilities and observed outcomes; a lower HL statistic indicates better calibration(Figure 4 ). DeLong Test The DeLong test was applied to both training and testing sets to evaluate the statistical significance of differences between models(Figure 5 ). Our combined model demonstrated a significant improvement over clinical models and imaging models. The combined model demonstrated a significant improvement over the clinical model and imaging model. However, the improvement over the combined model was not as pronounced. This could be attributed to the limited incremental information gain provided by the fusion of clinical data with the imaging model. This could be attributed to the limited incremental information gain provided by the fusion of clinical data with the imaging model. Decision Curve Analysis (DCA) Figure 6 presents the decision curve analysis (DCA) for both training and testing sets. The results indicate that our fusion model provides considerable advantages in terms of predicted probabilities. The fusion model provides significant advantages in terms of predicted probabilities. Furthermore, it consistently offers a greater net benefit compared to other signatures, underscoring its effectiveness. Discussion In this study, we performed a retrospective analysis and developed a combined model to identify high-risk individuals susceptible to osteoporosis in patients with COPD. In addition, we compared the advantages of the combined model with the clinical model and the radiomics model. Compared with a single model, the combined model has higher calibration and discrimination, and its application may help in the early identification of patients at risk for osteoporosis and early intervention, thereby reducing fracture risk and improving the overall net benefit for patients with COPD. It is important to assess COPD patients for comorbid osteoporosis because osteoporosis increases the risk of fracture. Adas-Okuma MG et al. showed that COPD is an independent risk factor for osteoporosis and fracture [ 21 ]; in addition, the risk of developing osteoporosis increases as the degree of COPD patients increases [ 22 ], and also showed that COPD patients have a The overall prevalence of osteoporosis was 38% compared to 15% in non-COPD patients [ 23 ], and in another study, the number of osteoporotic fractures increased in COPD patients, and the quality of life of the patients was proportionately reduced [ 24 ], as well as the fact that having osteoporosis reduces the functional exercise capacity of COPD patients who undergo a pulmonary rehabilitation program, which reduces the likelihood of improving their condition [ 25 ]. These studies emphasize the importance of early identification of COPD patients with comorbid osteoporosis, especially those with acute-phase exacerbations of COPD. The present study is a retrospective analysis that provides an effective way of early screening of COPD patients at risk of comorbid osteoporosis using chest CT imaging. Currently, the prediction of osteoporosis focuses on clinical data, quantitative imaging parameters, and radiomics.Bin Zhang et al. predicted osteoporosis with an AUC of 0.802 by performing radiomics feature extraction on lumbar spine X-rays and combining it with deep learning techniques [ 26 ]. Tao Zhen et al. used multiparameter lumbar spine magnetic resonance to establish an imageomics model, thus predicting osteoporosis [ 27 ]. Kaibin Fang et al. isolated thoracic vertebrae from conventional chest CT images and used imageomics with 3D deep learning techniques to predict osteoporosis, which resulted in an AUC of 0.906 [ 28 ]. Jing Pan et al. used a multi-feature deep learning model based on thoracic CT's multi-feature deep learning model, combined with clinical information and radiomics to screen for osteoporosis, with an AUC of 0.989 [ 29 ]. Currently, the majority of studies focus solely on predicting the risk of osteoporosis in healthy individuals using radiomics, with limited research establishing a connection between COPD and osteoporosis, Heqi Yang et al. developed a fracture risk prediction model for COPD patients by radiomics feature extraction of thoracic vertebrae from chest CT, combined with clinical data such as age and HDL level, with an AUC of 0.773 [ 30 ]. Furthermore, these studies confirmed the reliability of radiographic features extracted from bone structure in predicting osteoporosis outcomes. In contrast, the present study was based on the unique characteristics of COPD patients, and by extracting whole lung parenchyma imaging histologic features as well as clinical indicators, using univariate as well as multivariate analyses to incorporate features with P < 0.05, a clinical model, an imaging model, and a combined model were established, and the combined model had the highest efficacy in comparison with the combined model, which had an AUC of 0.811. It has been reported that constructing a column-line diagram by extracting whole-lung imaging histologic features for COPD patients has been used in several directions. Zhu Z et al. constructed a fusion model of deep learning and radiographic features by extracting whole-lung parenchymal imaging histologic features using a multilayer perceptron and then incorporated epidemiologic questionnaire data to build a more comprehensive early prediction model for COPD, which achieved an AUC of 0.971 [ 31 ]. Xiaoqing Lin et al. similarly constructed a column-line diagram by extracting whole-lung radiomics features and combining them with clinical features to identify cardiovascular risk in patients with COPD, and the model achieved good efficacy [ 16 ]. To date, there have been no studies identifying osteoporosis risk in COPD patients by whole-lung radiomics. Our study demonstrated the feasibility of this approach, although its performance was slightly inferior to that of a model constructed using features extracted from the bone structure itself. In the future, we will enhance the research on other related factors and integrate them with the existing models to improve the management efficiency for patients with COPD. Limitations of This Study There are some limitations to this study. First, due to the retrospective nature of this study, potential confounders and bias may be present, while some laboratory indicators were excluded because of missing values > 30%. Second, the categorization of COPD lung function indicators that might improve the validity of the model was not performed due to the imbalance of each classification. Future studies should have larger sample sizes and balanced classifications. Third, this study only examined lung characteristics to determine osteoporosis risk in patients with COPD and did not assess the validity of bone characteristics; future studies should include models based on bone characteristics and compare them to lung-based models. Fourth, future prospective studies should incorporate additional clinical factors as well as genetic characterization, particularly those highly associated with osteoporosis, to further assess the validity of radiographic scoring systems. Conclusion In this study, a new osteoporosis prediction model was constructed using automatic segmentation of whole lung parenchyma from CT examinations of patients with COPD and extraction of imaging histologic features, combined with other clinical variables, and compared with a single clinical model and imaging model. The results showed that the combined model had better results in predicting osteoporosis risk, which is expected to provide clinicians with a more accurate and practical tool for assessing osteoporosis risk and to give additional value to chest CT images of COPD patients. Meanwhile, this study also validated the association between osteoporosis and COPD. Future studies will focus on expanding the sample size, integrating quantitative features, and utilizing more advanced deep-learning techniques to optimize the model further. Declarations Funding: This work was supported by the Science and Technology Project of Huzhou City, Zhejiang Province (2023GY33) Acknowledgments : We sincerely thank Platform Onekey AI for Code consultation of the study. Competing interests : The author(s) declare no competing interests. Data availability : All data generated or analyzed during this study are included in this published article. Author Contribution H.P.B .and S.Q.Z.wrote the main text of the manuscript. J.N.X., L.Y.Q., and H.Y.Q. performed data processing. X.H.P., M.L., Y.F.Z., and P.L.X. were responsible for data collection and organization. H.X.Z. reviewed and edited the manuscript. All authors contributed to the overall review of the manuscript. Data Availability All data generated or analyzed during this study are included inthis published article. References Halpin DMG. Mortality of patients with COPD. Expert Rev Respir Med. 2024 Jun;18(6):381-395. Kahnert K, Jörres RA, Behr J, Welte T. The Diagnosis and Treatment of COPD and Its Comorbidities. Dtsch Arztebl Int. 2023 Jun 23;120(25):434-444. vespasiani-Gentilucci U, Pedone C, Muley-Vilamu M, et al. The pharmacological treatment of chronic comorbidities in COPD: mind the gap! Pulm Pharmacol Ther 2018; 51:48-58. Garvey C, Criner GJ. Impact of comorbidities on the treatment of chronic obstructive pulmonary disease. Am J Med 2018; 131(9S):23-29. ställberg B, Janson C, Larsson K, et al. Real-world retrospective cohort study ARCTIC shows the burden of comorbidities in Swedish COPD versus non-COPD patients. Prim Care Respir Med 2018; 28(1):33. gazzotti M, Roco C, Pradella C, et al. Frequency of osteoporosis and vertebral fractures in chronic obstructive pulmonary disease (COPD) patients. Arch Bronconeumol 2019; 55(5):252-257. m. Brennan, d. nash, r. rutherford. 87 vertebral fractures in older patients with COPD: an under-diagnosed and under-treated entity age aging. 51. Supple3. 2022. Dou Z, Chen X, Chen J, Yang H, Chen J. Chronic Obstructive Pulmonary Disease and Osteoporosis: a Two-Sample Mendelian Randomization Analysis. Chronic Obstr Pulm Dis. 2024 Jul 25;11(4):416-426. Shen L, Lv J, Li J, Zhou J, Wang X. Managing Osteoporosis in COPD. Endocr Metab Immune Disord Drug Targets. 2024;24(8):896-901. Fogelman l, Blake GM. Different approaches to bone densitometry. JNucl Med 2000; 41(12):2015-2025. Peng T, Zeng X, Li Y, Li M, Pu B, Zhi B, Wang Y, Qu H. A study on whether deep learning models based on CT images for bone density classification and prediction can be used for opportunistic osteoporosis screening. Osteoporos Int. 2024 Jan;35(1):117-128. Lambin P Leijenaar RTH, DeistTM et al. (2017) Radiomics: the bridge between medical imaging and personalized medicine. Nat Rev ClinOncol 14(12):749- 762. Mayerhoefer ME, Materka A, Langs G et al (2020) Introduction to radi-omics.JNucl Med 61(4):488-495. Zhu Z, Zhao S, Li J, Wang Y, Xu L, Jia Y, Li Z, Li W, Chen G, Wu X. Development and application of a deep learning-based comprehensive early diagnostic model for chronic obstructive pulmonary disease. Respir Res. 2024 Apr 18;25(1):167. Makimoto K, Hogg JC, Bourbeau J, Tan WC, Kirby M; CanCOLD Collaborative Research Group. Enhancing COPD classification using combined quantitative computed tomography and texture-based radiomics: a CanCOLD cohort study. ERJ Open Res. 2024 Jul 22;10(4):00968-2023. Lin X, Zhou T, Ni J, Li J, Guan Y, Jiang X, Zhou X, Xia Y, Xu F, Hu H, Dong Q, Liu S, Fan L. CT-based whole lung radiomics nomogram: a tool for identifying the risk of cardiovascular disease in patients with chronic obstructive pulmonary disease. Eur Radiol. 2024 Aug;34(8):4852-4863. Lin X, Shen R, Zheng X, Shi S, Dai Z, Fang K. Utilizing radiomics techniques to isolate a single vertebral body from chest CT for opportunistic osteoporosis screening. BMC Musculoskelet Disord. 2024 Oct 4;25(1):785. Labaki WW, Rosenberg SR. Chronic Obstructive Pulmonary Disease. Ann Intern Med. 2020 Aug 4;173(3): ITC17-ITC32. Halpin DMG, Criner GJ, Papi A, Singh D, Anzueto A, Martinez FJ, Agusti AA, Vogelmeier CF. Global Initiative for the Diagnosis, Management, and Prevention of Chronic Obstructive Lung Disease. The 2020 GOLD Science Committee Report on COVID-19 and Chronic Obstructive Pulmonary Disease. Am J Respir Crit Care Med. 2021 Jan 1;203(1):24-36. Brown KH, Ghita-Pettigrew M, Kerr BN, Mohamed-Smith L, Walls GM, McGarry CK, Butterworth KT. Characterization of quantitative imaging biomarkers for inflammatory and fibrotic radiation-induced lung injuries using preclinical radiomics. Radiother Oncol. 2024 Mar;192:110106. Adas-Okuma MG, Maeda SS, Gazzotti MR, Roco CM, Pradella CO, Nascimento OA, Porto EF, Vieira JGH, Jardim JR, Lazaretti-Castro M. COPD as an independent risk factor for osteoporosis and fractures. Osteoporos Int. 2020 Apr;31(4):687-697. Bitar AN, Sulaiman SAS, Ali IABH, Khan AH. Prevalence, risk assessment, and predictors of osteoporosis among chronic obstructive pulmonary disease patients. J Adv Pharm Technol Res. 2021 Oct-Dec;12(4):395-401. Chen YW, Ramsook AH, Coxson HO, et al. Prevalence and risk factors for osteoporosis in individuals with COPD: a systematic review and meta-analysis. Chest. 2019;156(6):1092-1110. Borgström F, Karlsson L, Ortsäter G, et al. Fragility fractures in Europe: burden, management, and opportunities. Arch Osteoporos. 2020;15(1):59. Li Y, Gao H, Zhao L, Wang J. Osteoporosis in COPD patients: risk factors and pulmonary rehabilitation. Clin Respir J. 2022 Jul;16(7):487-496. Zhang B, Chen Z, Yan R, Lai B, Wu G, You J, Wu X, Duan J, Zhang S. Development and Validation of a Feature-Based Broad-Learning System for Opportunistic Osteoporosis Screening Using Lumbar Spine Radiographs. Acad Radiol. 2024 Jan;31(1):84-92. Zhen T, Fang J, Hu D, Shen Q, Ruan M. Comparative evaluation of multiparametric lumbar MRI radiomic models for detecting osteoporosis. BMC Musculoskelet Disord. 2024 Feb 29;25(1):185. Fang K, Zheng X, Lin X, Dai Z. A comprehensive approach for osteoporosis detection through chest CT analysis and bone turnover markers: harnessing radiomics and deep learning techniques. Front Endocrinol (Lausanne). 2024 Jun 4;15:1296047. Pan J, Lin PC, Gong SC, Wang Z, Cao R, Lv Y, Zhang K, Wang L. Feasibility study of opportunistic osteoporosis screening on chest CT using a multi-feature fusion DCNN model. Arch Osteoporos. 2024 Oct 17;19(1):98. Yang H, Li Y, Yang H, Shi Z, Yao Q, Jia C, Song M, Qin J. A Novel CT-Based Fracture Risk Prediction Model for COPD Patients. Acad Radiol. 2024 Oct 10:S1076- 6332(24)00602-0. Zhu Z, Zhao S, Li J, Wang Y, Xu L, Jia Y, Li Z, Li W, Chen G, Wu X. Development and application of a deep learning-based comprehensive early diagnostic model for chronic obstructive pulmonary disease. Respir Res. 2024 Apr 18;25(1):167. Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterial.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6456138","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":456232286,"identity":"07c79d8c-a740-4321-9fff-289d4c0e37ac","order_by":0,"name":"Hupo Bian","email":"","orcid":"","institution":"The First Affiliated Hospital of Huzhou University","correspondingAuthor":false,"prefix":"","firstName":"Hupo","middleName":"","lastName":"Bian","suffix":""},{"id":456232287,"identity":"7a3d172f-e52a-45ff-8eba-0384a751e861","order_by":1,"name":"Shaoqi Zhu","email":"","orcid":"","institution":"The First Affiliated Hospital of Huzhou University","correspondingAuthor":false,"prefix":"","firstName":"Shaoqi","middleName":"","lastName":"Zhu","suffix":""},{"id":456232290,"identity":"07efc67e-fdd4-422c-8d33-cff55b4565a6","order_by":2,"name":"Huiying Qian","email":"","orcid":"","institution":"Huzhou Central Hospital affiliated to Huzhou University","correspondingAuthor":false,"prefix":"","firstName":"Huiying","middleName":"","lastName":"Qian","suffix":""},{"id":456232291,"identity":"c9e92a52-1d17-47a9-80be-f17f68729c2e","order_by":3,"name":"Jingnan Xue","email":"","orcid":"","institution":"The First Affiliated Hospital of Huzhou University","correspondingAuthor":false,"prefix":"","firstName":"Jingnan","middleName":"","lastName":"Xue","suffix":""},{"id":456232296,"identity":"8c472c58-8314-4c17-9986-b3b92db0a54c","order_by":4,"name":"Luying Qi","email":"","orcid":"","institution":"The First Affiliated Hospital of Huzhou University","correspondingAuthor":false,"prefix":"","firstName":"Luying","middleName":"","lastName":"Qi","suffix":""},{"id":456232301,"identity":"fd098da6-7a58-40ac-9fef-16c32a8ad069","order_by":5,"name":"Xiuhua Peng","email":"","orcid":"","institution":"The First Affiliated Hospital of Huzhou University","correspondingAuthor":false,"prefix":"","firstName":"Xiuhua","middleName":"","lastName":"Peng","suffix":""},{"id":456232303,"identity":"e3320303-8566-4aa7-b338-38c22d4ea562","order_by":6,"name":"Mei Li","email":"","orcid":"","institution":"The First Affiliated Hospital of Huzhou University","correspondingAuthor":false,"prefix":"","firstName":"Mei","middleName":"","lastName":"Li","suffix":""},{"id":456232304,"identity":"ae83fc6e-4204-4b59-919b-9287fa80041f","order_by":7,"name":"Yifeng Zheng","email":"","orcid":"","institution":"Huzhou Central Hospital affiliated to Huzhou University","correspondingAuthor":false,"prefix":"","firstName":"Yifeng","middleName":"","lastName":"Zheng","suffix":""},{"id":456232305,"identity":"0896a1bb-5f4e-46e7-99b8-3f880168ee12","order_by":8,"name":"Pengliang Xu","email":"","orcid":"","institution":"The First Affiliated Hospital of Huzhou University","correspondingAuthor":false,"prefix":"","firstName":"Pengliang","middleName":"","lastName":"Xu","suffix":""},{"id":456232309,"identity":"5d51d272-6faa-42d2-8c10-df5a64228c74","order_by":9,"name":"Hongxing Zhao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0klEQVRIiWNgGAWjYBAC+2bG9p8fKhiYDUA8HmK0GLA3H5CWOEOSFp5jCRK8bUAG0VrMJXIMDCTn1bGbSyQwPnjbxiBvTkiL5Ywcg4TCbWzMljMSmA3ntjEY7mwgpOdGjsEByW08zAY3EtikgS5MMDhAWIthA+8cCZAW9t9EaTE4cyyZgbfBAGwLM1FaJNubjzFLHEtgNjjzsFlyzjkJww2EtPAzM7YxfqipSzY4nnzww5syG3nCfoGCZAYGxgYgLUGkeiCwI17pKBgFo2AUjDgAAEnxO+o1DbsMAAAAAElFTkSuQmCC","orcid":"","institution":"The First Affiliated Hospital of Huzhou University","correspondingAuthor":true,"prefix":"","firstName":"Hongxing","middleName":"","lastName":"Zhao","suffix":""}],"badges":[],"createdAt":"2025-04-15 15:08:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6456138/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6456138/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82899693,"identity":"262ddc1b-aa2e-412b-923c-3d306d59e575","added_by":"auto","created_at":"2025-05-16 13:21:05","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":113368,"visible":true,"origin":"","legend":"\u003cp\u003eWorkflow of this study.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6456138/v1/120bd76a990938cc780cfc18.jpeg"},{"id":82899692,"identity":"4e0e5f8c-f177-4cae-97d0-f987a52efa34","added_by":"auto","created_at":"2025-05-16 13:21:05","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":108633,"visible":true,"origin":"","legend":"\u003cp\u003eOptimal hyperparameter λ was chosen through LASSO regression by 10-fold cross-validation, and the lowest value indicated the feature best matched the real observations (A); Radiomics features of non-zero coefficients were identified by LASSO regression models (B); 6 radiomics features together with associated coefficients following dimensionality reduction via LASSO regression (C).\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6456138/v1/2d2a778273a8974e43bd3470.jpeg"},{"id":82899697,"identity":"c43fb9e4-753c-4ee9-9f80-5242df24a06e","added_by":"auto","created_at":"2025-05-16 13:21:05","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":39574,"visible":true,"origin":"","legend":"\u003cp\u003eDifferent signatures' ROC on different cohorts.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6456138/v1/4a1a0a8d4dfd88ca9a63c207.jpeg"},{"id":82899698,"identity":"25fa5fb5-1a7d-4980-beca-64b264b9d05d","added_by":"auto","created_at":"2025-05-16 13:21:05","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":76129,"visible":true,"origin":"","legend":"\u003cp\u003eThe calibration curve predicts the OP of COPD patients (a: training cohort; b: internal validation cohort; c: external validation cohort). The developed radiomics nomogram to predict OP of COPD patients (d).\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6456138/v1/7ab1d6723ee641136076b85a.jpeg"},{"id":82901262,"identity":"1f540899-698a-4708-a267-e5b464795b94","added_by":"auto","created_at":"2025-05-16 13:29:05","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":40483,"visible":true,"origin":"","legend":"\u003cp\u003eDeLong test results for different signatures.\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6456138/v1/a35384b468a3f29bf54b5bb6.jpeg"},{"id":82899696,"identity":"77949303-9a7d-4d66-9f00-b8c527882759","added_by":"auto","created_at":"2025-05-16 13:21:05","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":59389,"visible":true,"origin":"","legend":"\u003cp\u003eThe decision curve analysis for the three models.\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6456138/v1/05eba2aa77653e97f63c1253.jpeg"},{"id":100948287,"identity":"91c17120-adb8-4b8d-b944-4a8f8808b6fb","added_by":"auto","created_at":"2026-01-23 06:41:42","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1663673,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6456138/v1/47411fe1-a297-4145-8530-277ec356f3fd.pdf"},{"id":82901261,"identity":"c8bee249-1d23-4c1d-9f2b-d325f5a26ccc","added_by":"auto","created_at":"2025-05-16 13:29:05","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":18171,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-6456138/v1/75d105e44c09c8cfa0967417.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"The potential value of CT-based whole lung radiomics nomogram for predicting osteoporosis risk in COPD patients: a two-center study","fulltext":[{"header":"Introduction","content":"\u003cp\u003eChronic Obstructive Pulmonary Disease (COPD) is one of the leading causes of death worldwide [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e], characterized by chronic airflow limitation, and in addition to its respiratory effects, COPD is often comorbid with other chronic diseases, including cardiovascular disease, osteoporosis, diabetes, lung cancer, cachexia, anemia, anxiety and depression [\u003cspan additionalcitationids=\"CR3 CR4\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Among these comorbidities, osteoporosis (OP) has emerged as an important comorbidity of COPD, increasing the risk of fracture, increasing patient morbidity and mortality [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], and causing a significant social burden.\u003c/p\u003e \u003cp\u003eAccording to the literature, patients with COPD are more likely to develop osteoporosis than those without COPD [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], and possible factors contributing to osteoporosis in patients with COPD include systemic inflammation, use of corticosteroids, vitamin D deficiency, smoking, hypoxia, and anemia [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. However, the link between COPD and osteoporosis is not fully understood. Bone mineral density (BMD) is widely recognized as the gold standard for the diagnosis of osteoporosis [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Still, clinicians do not routinely request BMD assessment in patients with COPD, and in the absence of BMD screening, CT scans are nevertheless widely used in patients with COPD. Although CT scans are not commonly used for the diagnosis of osteoporosis, opportunistic CT testing may improve the overall screening rate for osteoporosis [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] and may reduce the incidence of future fractures; therefore, it is critical to determine how to utilize CT testing in COPD patients to assess their risk of developing osteoporosis.\u003c/p\u003e \u003cp\u003eradiomics is an emerging high-throughput method for extracting quantitative image features [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] and has been successfully used in various aspects of early diagnosis [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], staging [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], and prediction of cardiovascular disease risk [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] in COPD patients. However, no investigator has yet utilized whole-lung radiomics to identify osteoporosis in COPD patients. Therefore, to meet clinical needs, it is necessary to investigate quantitative data obtained from lung parenchyma to identify the risk of osteoporosis in COPD patients. Based on this hypothesis, we aimed to evaluate the value of CT-based whole-lung radiomics features in determining the risk of osteoporosis in patients with COPD.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatient characteristics\u003c/h2\u003e \u003cp\u003ePatients and clinical data This retrospective study was approved by the ethics committees of both hospitals (No. 2025KYLL036-01) and waived the requirement for written informed consent. The study included 258 patients with COPD diagnosed by pulmonary function tests (PFT) from January 2020 to October 2024 at both centers. Inclusion criteria were as follows:(1) COPD diagnosed by PFT; (2) PFT and chest CT completed within 2 weeks; and (3) complete thin-layer (1-mm) chest CT images. Exclusion criteria were as follows:(1) combination of other chest diseases (e.g., pneumonia, pulmonary atelectasis, pulmonary nodules\u0026thinsp;\u0026gt;\u0026thinsp;6 mm or masses and pleural effusion; (2) combination of malignant tumors; and (3) spinal implants or significant imaging changes. A total of 258 patients were included for these etiologies. To make the study more rigorous and to prevent model overfitting, 213 patients from Hospital 1 were randomly assigned to either the training cohort (n\u0026thinsp;=\u0026thinsp;149) or the internal validation cohort (N\u0026thinsp;=\u0026thinsp;64) in a 7:3 ratio, and 45 patients from Hospital 2 were assigned to the external validation cohort (n\u0026thinsp;=\u0026thinsp;45). The training cohort was used to develop the model, while the internal and external validation cohorts were used to assess the validity of the model in clinical practice and to enhance the robustness of the study. Clinical information included age, weight, height, BMI, gender, GOLD class, smoking status, and laboratory tests such as C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), procalcitonin (PCT), albumin, alkaline phosphatase, triglycerides, platelet distribution width (PDW), arterial partial pressure of oxygen (Pa02), arterial partial pressure of carbon dioxide ( PACO2), white blood cell count, percent neutrophils, percent lymphocytes, absolute eosinophil count, erythrocyte pressure volume, erythrocyte distribution width, mean platelet volume, hemoglobin, and globulin. Laboratory factors with missing values greater than 30% were excluded due to their retrospective nature. Finally, 14 laboratory factors: C-reactive protein, albumin, Globulin, triglyceride, Alkaline Phosphatase, White blood cell count, Neutrophilic granulocyte Percentage, Percentage of lymphocytes, Absolute eosinophil count, Plateletcrit, Red blood cell distribution width, Mean platelet volume, Platelet Distribution Width, Hemoglobin were included in further analysis. Osteoporotic events were obtained by searching for GMD examinations. The presence or absence of osteoporotic events was diagnosed at the time of admission, and the interval between admission and chest CT scan was less than 1 month.The gold standard for the precise diagnosis of osteoporosis is the dual-energy X-ray absorptiometry (DXA) technique, in which a T-score of -2.5 or lower indicates the presence of osteoporosis[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eCT image acquisition and lung function testing\u003c/h3\u003e\n\u003cp\u003eParticipants underwent non-contrast CT scanning with Aquilion ONE TSX-301C, Somatom Force, Brillince CT 16, et al. Axial CT images of the entire chest were acquired under full inspiration (The scanning parameters are shown in Supplementary Material 1). Pulmonary function tests were performed using a CHEST multifunction spirometer HI-801 (Japan). Forced expiratory volume in 1 s (FEV1), percent predicted of forced expiratory flow in 1 s (FEV1%), and FEV1/FVC are the diagnostic criteria for COPD. The criterion is FEV1/FVC\u0026thinsp;\u0026lt;\u0026thinsp;0.7, and FEV1 increases less than 200 mL after the use of a bronchodilator [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Subjects were grouped according to the global initiative for chronic obstructive lung disease [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]: GOLD I grade, FEV1/FVC\u0026thinsp;\u0026lt;\u0026thinsp;0.7, FEV1\u0026thinsp;\u0026ge;\u0026thinsp;80% predicted; GOLD II grade, FEV1/FVC\u0026thinsp;\u0026lt;\u0026thinsp;0.7, 50% predicted\u0026thinsp;\u0026lt;\u0026thinsp;FEV1\u0026thinsp;\u0026le;\u0026thinsp;80% predicted; GOLD III grade, FEV1/FVC\u0026thinsp;\u0026lt;\u0026thinsp;0.7, 30% predicted\u0026thinsp;\u0026lt;\u0026thinsp;FEV1\u0026thinsp;\u0026le;\u0026thinsp;50% predicted; GOLD IV grade, FEVI/FVC\u0026thinsp;\u0026lt;\u0026thinsp;0.7, FEV1\u0026thinsp;\u0026lt;\u0026thinsp;30% predicted.\u003c/p\u003e\n\u003ch3\u003eFully automated region of interest segmentation\u003c/h3\u003e\n\u003cp\u003eWe automatically segmented the right and left lungs by the Full Lung Segmentation component on the OnekeyAI platform (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/OnekeyAI-Platform/onekey\u003c/span\u003e\u003cspan address=\"https://github.com/OnekeyAI-Platform/onekey\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), and the extracted left and right lungs were merged into a region of interest (ROI) (the specific algorithm process can be found in Supplementary Material 2). The segmentation results were then independently assessed by two chest radiologists with more than 10 years of experience using ITK-SNAP software. [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] (version 3.8.0, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.itksnap.org\u003c/span\u003e\u003cspan address=\"http://www.itksnap.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to correct erroneous segmentations.\u003c/p\u003e\n\u003ch3\u003eWhole lung radiomics feature extraction\u003c/h3\u003e\n\u003cp\u003eRadiomic feature extraction was conducted using the PyRadiomics open-source Python package (version 3.7.12; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pyradiomics.readthedocs.io\u003c/span\u003e\u003cspan address=\"https://pyradiomics.readthedocs.io\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) on the OnekeyAI platform, with three features obtained: first-order features, shape, and texture feature. Radiation features were extracted using Z-score normalization. All available features implemented in PyRadiomics were extracted from the original images and filtered images, including wavelet and Laplacian of Gaussian transformations. Before feature extraction, the image underwent a three-step preprocessing operation, voxel resampling, gray discretization, and image intensity normalization, resampling the image to 1 mm *1 mm * 1 mm, and adjusting the gray value of the image to 25 gray levels, which is helpful to reduce the influence of different layer thicknesses and reduce the interference of noise.\u003c/p\u003e\n\u003ch3\u003eRadiomics feature selection and model construction\u003c/h3\u003e\n\u003cp\u003eWe normalized all extracted features using Z-score normalization and performed a t-test to assess their statistical significance, retaining only those with a p-value below 0.05. To mitigate collinearity, we examined the correlations between features using Pearson's correlation coefficient, excluding any feature from pairs that exhibited a correlation coefficient above 0.9. We further refined the feature set through Lasso regression within a 10-fold cross-validation framework to identify the optimal regularization parameter, λ, effectively reducing the feature set to those most predictive and informative.\u003c/p\u003e \u003cp\u003eRadiomics Signature: By employing LASSO for stringent feature selection, we constructed the radiomics risk model using machine learning algorithms such as Logistic Regression, Support Vector Machine, and Random Forest. Comparative analyses were performed to gauge each model's performance, and we investigated the benefits of feature fusion by integrating multi-modal features, allowing us to explore the advantages of combining multiple imaging modalities to enhance predictive accuracy.\u003c/p\u003e \u003cp\u003eThe diagnostic efficacy of our machine learning model was rigorously evaluated in the test cohort through the construction of Receiver Operating Characteristic (ROC) curves to assess its discriminative ability,furthermore, for all models, 5-fold cross-validation was conducted. The calibration of the model was analyzed using calibration curves, complemented by the Hosmer-Lemeshow goodness-of-fit test to validate its reliability. Additionally, Decision Curve Analysis (DCA) was employed to appraise the clinical utility of our predictive models, facilitating an understanding of the potential benefits in a clinical context.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eMetrics\u003c/h2\u003e \u003cp\u003eThe diagnostic efficacy of our machine learning model was rigorously evaluated in the test cohort through the construction of Receiver Operating Characteristic (ROC) curves to assess its discriminative ability. The calibration of the model was analyzed using calibration curves, complemented by the Hosmer-Lemeshow goodness-of-fit test to validate its reliability. Additionally, Decision Curve Analysis (DCA) was employed to appraise the clinical utility of our predictive models, facilitating an understanding of the potential benefits in a clinical context.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eThe normality of clinical features was verified using the Shapiro-Wilk test. Continuous variables were subjected to the t-test or the Mann-Whitney U test based on their distribution characteristics. Categorical variables were examined using Chi-square (χ\u0026sup2;) tests. The baseline characteristics across all cohorts are detailed in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. Notably, p-values exceeding 0.05 between cohorts indicated no significant differences, thus confirming an unbiased division between groups.\u003c/p\u003e \u003cp\u003eAll data analyses were executed on the OnekeyAI platform version 4.9.1 utilizing Python 3.7.12. Statistical assessments were conducted with Statsmodels version 0.13.2. Radiomics feature extraction was performed via PyRadiomics version 3.0.1. Machine learning implementations, including the Support Vector Machine (SVM), were facilitated using Scikit-learn version 1.0.2.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eThe flow chart for patient selection is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. As of October 31, 2024, we included a total of 434 patients with a diagnosis of COPD, and after screening by exclusion criteria, the final cohort (n\u0026thinsp;=\u0026thinsp;258) included 80 women and 178 men, with a mean age of 74.98 years for the entire cohort, and the two hospitals included 213 and 45 patients, respectively. Among them, 72 COPD patients did not have osteoporosis, while 186 COPD patients had osteoporosis.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eModeling of Clinical Factors\u003c/h2\u003e \u003cp\u003eThe clinical factors of the patients in the training and test sets are shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Independent risk factor identification (univariate and multivariate logistic regression analysis) is shown in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. In the univariate analysis, height, hemoglobin, C-reactive protein (CRP), alkaline phosphatase (ALP), globulin, weight, age, neutrophilic granulocyte percentage, erythrocyte sedimentation rate (ESR), albumin, arterial carbon dioxide partial pressure, BMI, and other factors. Dioxide partial pressure, BMI, Vitamin D, platelet distribution width (PDW), procalcitonin, red blood cell distribution width (RDW), mean platelet volume (MPV), white blood cell distribution width (RDW), and mean platelet volume (MPV). Volume (MPV), white blood cell count (WBC), triglyceride levels, and smoking status were all associated with the outcome, as indicated by ORs significantly different from 1 and p-values\u0026thinsp;\u0026lt;\u0026thinsp;0.05 (or 0, indicating\u0026thinsp;\u0026lt;\u0026thinsp;0.001). However, in the multivariate analysis, only neutrophilic granulocyte percentage (OR\u0026thinsp;=\u0026thinsp;1.044, p\u0026thinsp;=\u0026thinsp;0.042) and platelet crit (OR\u0026thinsp;=\u0026thinsp;0.0, p\u0026thinsp;=\u0026thinsp;0.044) retained significance. The other features exhibited ORs close to 1 and higher p-values, suggesting that their associations with the outcome were not independent of the different variables considered in the model. of the different variables considered in the model. Finally, neutrophil percentage and platelet count were included in the construction of the clinical factors model.\u003c/p\u003e \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003ctable id=\"Tab1\" border=\"1\" class=\"fr-table-selection-hover\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eBaseline characteristics of the study population\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" rowspan=\"2\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eClinical factors\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" rowspan=\"2\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\" colspan=\"3\"\u003e\n \u003cp\u003eTraining cohort(n\u0026thinsp;=\u0026thinsp;149)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\" colspan=\"3\"\u003e\n \u003cp\u003eInternal validation cohort (n\u0026thinsp;=\u0026thinsp;64)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\" colspan=\"3\"\u003e\n \u003cp\u003eExternal vaidation cohort(n\u0026thinsp;=\u0026thinsp;45)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCOPD without OP(n\u0026thinsp;=\u0026thinsp;50)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCOPD with OP(n\u0026thinsp;=\u0026thinsp;99)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ep value\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCOPD without OP(n\u0026thinsp;=\u0026thinsp;17)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCOPD with OP(n\u0026thinsp;=\u0026thinsp;47)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ep value\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCOPD without OP(n\u0026thinsp;=\u0026thinsp;5)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eCOPD with OP(n\u0026thinsp;=\u0026thinsp;40)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ep value\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e72.14\u0026thinsp;\u0026plusmn;\u0026thinsp;10.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e76.20\u0026thinsp;\u0026plusmn;\u0026thinsp;8.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.042\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e71.53\u0026thinsp;\u0026plusmn;\u0026thinsp;7.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e77.81\u0026thinsp;\u0026plusmn;\u0026thinsp;8.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e68.00\u0026thinsp;\u0026plusmn;\u0026thinsp;10.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e74.53\u0026thinsp;\u0026plusmn;\u0026thinsp;9.67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.164\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eSmoke\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.66\u0026thinsp;\u0026plusmn;\u0026thinsp;0.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.53\u0026thinsp;\u0026plusmn;\u0026thinsp;0.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.118\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.65\u0026thinsp;\u0026plusmn;\u0026thinsp;0.49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.32\u0026thinsp;\u0026plusmn;\u0026thinsp;0.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.40\u0026thinsp;\u0026plusmn;\u0026thinsp;0.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.55\u0026thinsp;\u0026plusmn;\u0026thinsp;0.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.545\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eHeight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e168.26\u0026thinsp;\u0026plusmn;\u0026thinsp;4.96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e162.36\u0026thinsp;\u0026plusmn;\u0026thinsp;8.99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e168.71\u0026thinsp;\u0026plusmn;\u0026thinsp;6.66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e159.23\u0026thinsp;\u0026plusmn;\u0026thinsp;7.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e175.20\u0026thinsp;\u0026plusmn;\u0026thinsp;3.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e162.93\u0026thinsp;\u0026plusmn;\u0026thinsp;8.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eWeight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e66.66\u0026thinsp;\u0026plusmn;\u0026thinsp;10.40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e57.32\u0026thinsp;\u0026plusmn;\u0026thinsp;11.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e64.90\u0026thinsp;\u0026plusmn;\u0026thinsp;9.52\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e54.93\u0026thinsp;\u0026plusmn;\u0026thinsp;10.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e75.10\u0026thinsp;\u0026plusmn;\u0026thinsp;14.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e54.52\u0026thinsp;\u0026plusmn;\u0026thinsp;10.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eBMI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e23.55\u0026thinsp;\u0026plusmn;\u0026thinsp;3.68\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e21.68\u0026thinsp;\u0026plusmn;\u0026thinsp;3.69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e22.76\u0026thinsp;\u0026plusmn;\u0026thinsp;2.69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e21.62\u0026thinsp;\u0026plusmn;\u0026thinsp;3.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.224\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e24.41\u0026thinsp;\u0026plusmn;\u0026thinsp;4.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e20.51\u0026thinsp;\u0026plusmn;\u0026thinsp;3.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.017\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eC-Reactive Protein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e58.96\u0026thinsp;\u0026plusmn;\u0026thinsp;44.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e60.80\u0026thinsp;\u0026plusmn;\u0026thinsp;49.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.966\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e54.35\u0026thinsp;\u0026plusmn;\u0026thinsp;46.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e72.30\u0026thinsp;\u0026plusmn;\u0026thinsp;48.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.221\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e69.40\u0026thinsp;\u0026plusmn;\u0026thinsp;55.36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e103.42\u0026thinsp;\u0026plusmn;\u0026thinsp;55.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.155\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003ealbumin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e37.85\u0026thinsp;\u0026plusmn;\u0026thinsp;4.79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e37.80\u0026thinsp;\u0026plusmn;\u0026thinsp;4.14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.97\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e40.30\u0026thinsp;\u0026plusmn;\u0026thinsp;4.31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e37.25\u0026thinsp;\u0026plusmn;\u0026thinsp;4.91\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.027\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e37.46\u0026thinsp;\u0026plusmn;\u0026thinsp;5.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e35.54\u0026thinsp;\u0026plusmn;\u0026thinsp;4.57\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.386\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eGlobulin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e67.12\u0026thinsp;\u0026plusmn;\u0026thinsp;35.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e66.23\u0026thinsp;\u0026plusmn;\u0026thinsp;32.54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.915\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e70.24\u0026thinsp;\u0026plusmn;\u0026thinsp;36.52\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e63.09\u0026thinsp;\u0026plusmn;\u0026thinsp;36.96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.495\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e83.40\u0026thinsp;\u0026plusmn;\u0026thinsp;24.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e55.08\u0026thinsp;\u0026plusmn;\u0026thinsp;33.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.073\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003etriglyceride\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.32\u0026thinsp;\u0026plusmn;\u0026thinsp;0.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.04\u0026thinsp;\u0026plusmn;\u0026thinsp;0.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.35\u0026thinsp;\u0026plusmn;\u0026thinsp;0.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.17\u0026thinsp;\u0026plusmn;\u0026thinsp;0.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.294\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.82\u0026thinsp;\u0026plusmn;\u0026thinsp;0.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.97\u0026thinsp;\u0026plusmn;\u0026thinsp;0.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eAlkaline Phosphatase\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e86.88\u0026thinsp;\u0026plusmn;\u0026thinsp;67.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e87.72\u0026thinsp;\u0026plusmn;\u0026thinsp;36.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e84.29\u0026thinsp;\u0026plusmn;\u0026thinsp;21.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e89.94\u0026thinsp;\u0026plusmn;\u0026thinsp;42.25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.909\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e77.78\u0026thinsp;\u0026plusmn;\u0026thinsp;18.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e78.03\u0026thinsp;\u0026plusmn;\u0026thinsp;22.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.899\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eWhite blood cell count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6.47\u0026thinsp;\u0026plusmn;\u0026thinsp;2.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6.91\u0026thinsp;\u0026plusmn;\u0026thinsp;3.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.359\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6.96\u0026thinsp;\u0026plusmn;\u0026thinsp;2.88\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7.57\u0026thinsp;\u0026plusmn;\u0026thinsp;3.86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.508\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7.42\u0026thinsp;\u0026plusmn;\u0026thinsp;1.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7.62\u0026thinsp;\u0026plusmn;\u0026thinsp;2.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.866\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eNeutrophilic granulocyte percentage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e66.31\u0026thinsp;\u0026plusmn;\u0026thinsp;12.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e73.67\u0026thinsp;\u0026plusmn;\u0026thinsp;12.36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e67.34\u0026thinsp;\u0026plusmn;\u0026thinsp;11.81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e75.39\u0026thinsp;\u0026plusmn;\u0026thinsp;9.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e65.36\u0026thinsp;\u0026plusmn;\u0026thinsp;4.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e70.77\u0026thinsp;\u0026plusmn;\u0026thinsp;16.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.073\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003ePercentage of lymphocytes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e24.16\u0026thinsp;\u0026plusmn;\u0026thinsp;11.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17.60\u0026thinsp;\u0026plusmn;\u0026thinsp;8.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e23.61\u0026thinsp;\u0026plusmn;\u0026thinsp;9.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16.46\u0026thinsp;\u0026plusmn;\u0026thinsp;7.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e22.98\u0026thinsp;\u0026plusmn;\u0026thinsp;4.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18.26\u0026thinsp;\u0026plusmn;\u0026thinsp;12.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.093\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eAbsolute eosinophil count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.19\u0026thinsp;\u0026plusmn;\u0026thinsp;0.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.14\u0026thinsp;\u0026plusmn;\u0026thinsp;0.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.13\u0026thinsp;\u0026plusmn;\u0026thinsp;0.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.09\u0026thinsp;\u0026plusmn;\u0026thinsp;0.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.073\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.15\u0026thinsp;\u0026plusmn;\u0026thinsp;0.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.36\u0026thinsp;\u0026plusmn;\u0026thinsp;1.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.814\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003ePlateletcrit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.20\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.18\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.256\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.19\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.18\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.35\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.19\u0026thinsp;\u0026plusmn;\u0026thinsp;0.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.21\u0026thinsp;\u0026plusmn;\u0026thinsp;0.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.662\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eRed blood cell distribution width\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13.46\u0026thinsp;\u0026plusmn;\u0026thinsp;1.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13.49\u0026thinsp;\u0026plusmn;\u0026thinsp;1.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.904\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13.44\u0026thinsp;\u0026plusmn;\u0026thinsp;0.95\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13.75\u0026thinsp;\u0026plusmn;\u0026thinsp;1.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.873\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12.90\u0026thinsp;\u0026plusmn;\u0026thinsp;0.81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13.65\u0026thinsp;\u0026plusmn;\u0026thinsp;1.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.233\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eMean platelet volume\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10.10\u0026thinsp;\u0026plusmn;\u0026thinsp;1.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10.17\u0026thinsp;\u0026plusmn;\u0026thinsp;1.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.707\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10.56\u0026thinsp;\u0026plusmn;\u0026thinsp;2.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10.18\u0026thinsp;\u0026plusmn;\u0026thinsp;1.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.808\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10.58\u0026thinsp;\u0026plusmn;\u0026thinsp;1.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10.61\u0026thinsp;\u0026plusmn;\u0026thinsp;1.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.96\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003ePlatelet Distribution Width\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16.24\u0026thinsp;\u0026plusmn;\u0026thinsp;0.96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16.28\u0026thinsp;\u0026plusmn;\u0026thinsp;0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e15.89\u0026thinsp;\u0026plusmn;\u0026thinsp;1.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16.53\u0026thinsp;\u0026plusmn;\u0026thinsp;3.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.502\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e15.38\u0026thinsp;\u0026plusmn;\u0026thinsp;3.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12.37\u0026thinsp;\u0026plusmn;\u0026thinsp;2.82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.053\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eHemoglobin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e126.66\u0026thinsp;\u0026plusmn;\u0026thinsp;14.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e121.09\u0026thinsp;\u0026plusmn;\u0026thinsp;18.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e129.24\u0026thinsp;\u0026plusmn;\u0026thinsp;14.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e119.36\u0026thinsp;\u0026plusmn;\u0026thinsp;19.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.062\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e141.40\u0026thinsp;\u0026plusmn;\u0026thinsp;22.61\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e123.42\u0026thinsp;\u0026plusmn;\u0026thinsp;17.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eGOLD grade\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.251\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.393\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.132\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\n \u003cp\u003eGOLDⅠ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e23(46.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e37(37.37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7(41.18)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e14(29.79)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3(60.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7(17.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\n \u003cp\u003eGOLDⅡ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17(34.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e27(27.27)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4(23.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10(21.28)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1(20.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e8(20.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\n \u003cp\u003eGOLDⅢ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e8(16.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e19(19.19)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6(35.29)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18(38.30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e19(47.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\n \u003cp\u003eGOLDⅣ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2(4.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16(16.16)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5(10.64)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1(20.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6(15.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\n \u003cp\u003eGender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.371\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4(8.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e37(37.37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1(5.88)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e26(55.32)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12(30.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 11.6993%;\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\" style=\"width: 7.7778%;\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e46(92.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e62(62.63)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16(94.12)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e21(44.68)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e28(70.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eUnivariable and multivariable logistic regression analysis\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" rowspan=\"2\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eVariable\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"2\" style=\"width: 30.6376%;\"\u003e\n \u003cp\u003eUnivariable analysis\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"2\" style=\"width: 28.4603%;\"\u003e\n \u003cp\u003eMultivariable analysis\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003eOR[95%CI]\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003ep value\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003eOR[95%CI]\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003ep value\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003ePlateletcrit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e19.668[4.581,84.437]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e0[0,0.23]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.044\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eHemoglobin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.005[1.003,1.007]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e0.989[0.962,1.017]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.519\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eC-Reactive Protein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.007[1.004,1.011]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e0.996[0.988,1.004]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.44\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eAlkaline Phosphatase\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.007[1.003,1.01]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e0.997[0.989,1.005]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.512\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eGlobulin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.008[1.004,1.012]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.003[0.99,1.015]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.714\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eNeutrophilic granulocyte percentage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.011[1.007,1.015]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.044[1.008,1.08]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.042\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003ealbumin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.018[1.01,1.025]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.053[0.951,1.166]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.401\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003ePlatelet Distribution Width\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.043[1.025,1.062]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.229[0.768,1.966]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.471\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eRed blood cell distribution width\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.052[1.029,1.075]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e0.887[0.641,1.228]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.544\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eMean platelet volume\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.07[1.04,1.1]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e0.849[0.587,1.226]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.463\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eWhite blood cell count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.101[1.057,1.147]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.154[0.994,1.34]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.115\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003etriglyceride\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.325[1.068,1.642]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e0.031\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e0.696[0.412,1.175]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.255\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eGOLD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.415[1.232,1.624]]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.178[0.856,1.624]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eSmoke\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.576[1.093,2.273]]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e0.041\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.116[0.48,2.591]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.831\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eHeight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.004[1.002,1.006]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.001[0.923,1.084]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.986\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eWeight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.009[1.004,1.013]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e0.879[0.771,1.003]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.109\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.01[1.006,1.014]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.022[0.976,1.069]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.439\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" style=\"width: 37.9471%;\"\u003e\n \u003cp\u003eBMI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 22.5505%;\"\u003e\n \u003cp\u003e1.027[1.014,1.04]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.0871%;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 20.0622%;\"\u003e\n \u003cp\u003e1.264[0.884,1.806]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" style=\"width: 8.3981%;\"\u003e\n \u003cp\u003e0.281\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e \u003cp\u003e \u003cb\u003eFeature extraction\u003c/b\u003e, \u003cb\u003eselection, and radiomics signature construction\u003c/b\u003e\u003c/p\u003e \u003cp\u003eAfter extracting, normalizing, and deleting features with a correlation greater than 0.9 from CT images, a total of 107 radiological features were retained. We will select features using both mRMR and LASSO. First, mRMR was performed to eliminate redundant and irrelevant features, and 15 features were retained. Then, LASSO is performed to select an optimized subset of features to construct the final model (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea and b). Finally, six features were used to build the radiomics feature signature (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). Radiomics features were screened using the following formula. The rad score is shown as follows.\u003c/p\u003e \u003cp\u003eRadscore\u0026thinsp;=\u0026thinsp;0.6621621621621622 -0.095484 * original_shape_SurfaceArea\u0026thinsp;+\u0026thinsp;0.080137 * original_glszm_SizeZoneNonUniformityNormalized\u0026thinsp;+\u0026thinsp;0.016017 * original_glrlm_LongRunHighGrayLevelEmphasis\u0026thinsp;+\u0026thinsp;0.004569 * original_ngtdm_Strength \u0026minus;\u0026thinsp;0.030314 * original_glszm_SmallAreaLowGrayLevelEmphasis \u0026minus;\u0026thinsp;0.074244 * original_shape_Maximum3DDiameter\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eDevelopment of radiological column line diagrams and evaluation of the performance of different models\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e displays the performance of the clinical model, radiomics signature, and nomogram in the diagnosis of training, internal, and external validation sets. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents that the Clinic, Radiomics, and Combined indicators are presented across the training, validation, and test cohorts. In the training cohort, the Combined indicator yields the highest AUC of 0.811, followed by Radiomics (0.762) and Clinic (0.691). In the validation cohort, the combined indicator maintains a strong performance with an AUC of 0.806, while Radiomics achieves a slightly higher AUC of 0.765 compared to Clinic (0.724)c In the test cohort, the Combined indicator demonstrates an AUC of 0.728, marginally higher than Radiomics (0.718) and significantly higher than Clinic ( 0.656). Notably, the confidence intervals for the AUCs, particularly in the test cohort, are wide, indicating variability in performance.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMetrics on different signature\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSignature\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e95% CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSensitivity\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSpecificity\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003ePPV\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eNPV\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eCohort\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eClinic\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.649\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.691\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.5985\u0026ndash;0.7827\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.602\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.819\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.487\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRadiomics\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.762\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.6814\u0026ndash;0.8418\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.847\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.651\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCombined\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.696\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.811\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.7425\u0026ndash;0.8795\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.612\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.896\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003etrain\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eClinic\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.683\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.724\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.5881\u0026ndash;0.8594\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.824\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.452\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eval\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRadiomics\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.651\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.765\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.6427\u0026ndash;0.8867\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.543\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.941\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.962\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.432\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eval\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCombined\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.806\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.6873\u0026ndash;0.9239\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.696\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.824\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.914\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eval\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eClinic\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.659\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.656\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.4009\u0026ndash;0.9119\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.641\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.962\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.222\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003etest\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRadiomics\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.841\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.718\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.4065\u0026ndash;1.0000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.872\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.944\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.375\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003etest\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCombined\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.886\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.728\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.3582\u0026ndash;1.0000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.923\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.947\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003etest\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe analysis reveals that the Combined indicator consistently outperforms both the Clinic and Radiomics indicators in terms of AUC in the training and validation cohorts. Although the Combined indicator retains a higher AUC in the test cohort compared to the individual indicators, the wide confidence intervals suggest potential variability in its performance. Although the Combined indicator retains a higher AUC in the test cohort compared to the individual indicators, the wide confidence intervals suggest potential variability in its performance. The Radiomics indicator shows a notable increase in specificity across all cohorts, peaking at 0.941 in the validation cohort, but this is accompanied by lower sensitivity. The Clinic indicator demonstrates moderate performance but consistently exhibits lower AUC values and specificity compared to the other indicators. Overall, the Combined indicator appears to be the most robust across different cohorts, balancing the AUC values and specificity across different cohorts. Overall, the Combined indicator is the most robust across different cohorts, balancing both sensitivity and specificity, despite some variability in performance indicated by the wide confidence intervals. The Combined indicator appears to be the most robust across different cohorts, balancing both sensitivity and specificity, despite some variability in performance indicated by the wide confidence intervals.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eCalibration Curve\u003c/strong\u003e \u003cp\u003eThe Hosmer-Lemeshow (HL) test quantifies the discrepancy between predicted probabilities and observed outcomes; a lower HL statistic indicates better calibration(Figure \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eDeLong Test\u003c/strong\u003e \u003cp\u003eThe DeLong test was applied to both training and testing sets to evaluate the statistical significance of differences between models(Figure \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Our combined model demonstrated a significant improvement over clinical models and imaging models. The combined model demonstrated a significant improvement over the clinical model and imaging model. However, the improvement over the combined model was not as pronounced. This could be attributed to the limited incremental information gain provided by the fusion of clinical data with the imaging model. This could be attributed to the limited incremental information gain provided by the fusion of clinical data with the imaging model.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eDecision Curve Analysis (DCA)\u003c/strong\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e presents the decision curve analysis (DCA) for both training and testing sets. The results indicate that our fusion model provides considerable advantages in terms of predicted probabilities. The fusion model provides significant advantages in terms of predicted probabilities. Furthermore, it consistently offers a greater net benefit compared to other signatures, underscoring its effectiveness.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we performed a retrospective analysis and developed a combined model to identify high-risk individuals susceptible to osteoporosis in patients with COPD. In addition, we compared the advantages of the combined model with the clinical model and the radiomics model. Compared with a single model, the combined model has higher calibration and discrimination, and its application may help in the early identification of patients at risk for osteoporosis and early intervention, thereby reducing fracture risk and improving the overall net benefit for patients with COPD.\u003c/p\u003e \u003cp\u003eIt is important to assess COPD patients for comorbid osteoporosis because osteoporosis increases the risk of fracture. Adas-Okuma MG et al. showed that COPD is an independent risk factor for osteoporosis and fracture [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]; in addition, the risk of developing osteoporosis increases as the degree of COPD patients increases [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], and also showed that COPD patients have a The overall prevalence of osteoporosis was 38% compared to 15% in non-COPD patients [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e], and in another study, the number of osteoporotic fractures increased in COPD patients, and the quality of life of the patients was proportionately reduced [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e], as well as the fact that having osteoporosis reduces the functional exercise capacity of COPD patients who undergo a pulmonary rehabilitation program, which reduces the likelihood of improving their condition [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. These studies emphasize the importance of early identification of COPD patients with comorbid osteoporosis, especially those with acute-phase exacerbations of COPD. The present study is a retrospective analysis that provides an effective way of early screening of COPD patients at risk of comorbid osteoporosis using chest CT imaging.\u003c/p\u003e \u003cp\u003eCurrently, the prediction of osteoporosis focuses on clinical data, quantitative imaging parameters, and radiomics.Bin Zhang et al. predicted osteoporosis with an AUC of 0.802 by performing radiomics feature extraction on lumbar spine X-rays and combining it with deep learning techniques [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Tao Zhen et al. used multiparameter lumbar spine magnetic resonance to establish an imageomics model, thus predicting osteoporosis [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Kaibin Fang et al. isolated thoracic vertebrae from conventional chest CT images and used imageomics with 3D deep learning techniques to predict osteoporosis, which resulted in an AUC of 0.906 [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Jing Pan et al. used a multi-feature deep learning model based on thoracic CT's multi-feature deep learning model, combined with clinical information and radiomics to screen for osteoporosis, with an AUC of 0.989 [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Currently, the majority of studies focus solely on predicting the risk of osteoporosis in healthy individuals using radiomics, with limited research establishing a connection between COPD and osteoporosis, Heqi Yang et al. developed a fracture risk prediction model for COPD patients by radiomics feature extraction of thoracic vertebrae from chest CT, combined with clinical data such as age and HDL level, with an AUC of 0.773 [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Furthermore, these studies confirmed the reliability of radiographic features extracted from bone structure in predicting osteoporosis outcomes. In contrast, the present study was based on the unique characteristics of COPD patients, and by extracting whole lung parenchyma imaging histologic features as well as clinical indicators, using univariate as well as multivariate analyses to incorporate features with P\u0026thinsp;\u0026lt;\u0026thinsp;0.05, a clinical model, an imaging model, and a combined model were established, and the combined model had the highest efficacy in comparison with the combined model, which had an AUC of 0.811.\u003c/p\u003e \u003cp\u003eIt has been reported that constructing a column-line diagram by extracting whole-lung imaging histologic features for COPD patients has been used in several directions. Zhu Z et al. constructed a fusion model of deep learning and radiographic features by extracting whole-lung parenchymal imaging histologic features using a multilayer perceptron and then incorporated epidemiologic questionnaire data to build a more comprehensive early prediction model for COPD, which achieved an AUC of 0.971 [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Xiaoqing Lin et al. similarly constructed a column-line diagram by extracting whole-lung radiomics features and combining them with clinical features to identify cardiovascular risk in patients with COPD, and the model achieved good efficacy [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. To date, there have been no studies identifying osteoporosis risk in COPD patients by whole-lung radiomics. Our study demonstrated the feasibility of this approach, although its performance was slightly inferior to that of a model constructed using features extracted from the bone structure itself. In the future, we will enhance the research on other related factors and integrate them with the existing models to improve the management efficiency for patients with COPD.\u003c/p\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eLimitations of This Study\u003c/h2\u003e \u003cp\u003eThere are some limitations to this study. First, due to the retrospective nature of this study, potential confounders and bias may be present, while some laboratory indicators were excluded because of missing values\u0026thinsp;\u0026gt;\u0026thinsp;30%. Second, the categorization of COPD lung function indicators that might improve the validity of the model was not performed due to the imbalance of each classification. Future studies should have larger sample sizes and balanced classifications. Third, this study only examined lung characteristics to determine osteoporosis risk in patients with COPD and did not assess the validity of bone characteristics; future studies should include models based on bone characteristics and compare them to lung-based models. Fourth, future prospective studies should incorporate additional clinical factors as well as genetic characterization, particularly those highly associated with osteoporosis, to further assess the validity of radiographic scoring systems.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn this study, a new osteoporosis prediction model was constructed using automatic segmentation of whole lung parenchyma from CT examinations of patients with COPD and extraction of imaging histologic features, combined with other clinical variables, and compared with a single clinical model and imaging model. The results showed that the combined model had better results in predicting osteoporosis risk, which is expected to provide clinicians with a more accurate and practical tool for assessing osteoporosis risk and to give additional value to chest CT images of COPD patients. Meanwhile, this study also validated the association between osteoporosis and COPD. Future studies will focus on expanding the sample size, integrating quantitative features, and utilizing more advanced deep-learning techniques to optimize the model further.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding:\u003c/h2\u003e \u003cp\u003eThis work was supported by the Science and Technology Project of Huzhou City, Zhejiang Province (2023GY33)\u003c/p\u003e \u003cp\u003e \u003cb\u003eAcknowledgments\u003c/b\u003e: We sincerely thank Platform Onekey AI for Code consultation of the study.\u003c/p\u003e \u003cp\u003e \u003cb\u003eCompeting interests\u003c/b\u003e: The author(s) declare no competing interests.\u003c/p\u003e \u003cp\u003e \u003cb\u003eData availability\u003c/b\u003e: All data generated or analyzed during this study are included in\u003c/p\u003e \u003cp\u003ethis published article.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eH.P.B .and S.Q.Z.wrote the main text of the manuscript. J.N.X., L.Y.Q., and H.Y.Q. performed data processing. X.H.P., M.L., Y.F.Z., and P.L.X. were responsible for data collection and organization. H.X.Z. reviewed and edited the manuscript. All authors contributed to the overall review of the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eAll data generated or analyzed during this study are included inthis published article.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eHalpin DMG. Mortality of patients with COPD. Expert Rev Respir Med. 2024 Jun;18(6):381-395. \u003c/li\u003e\n\u003cli\u003eKahnert K, J\u0026ouml;rres RA, Behr J, Welte T. The Diagnosis and Treatment of COPD and Its Comorbidities. Dtsch Arztebl Int. 2023 Jun 23;120(25):434-444. \u003c/li\u003e\n\u003cli\u003evespasiani-Gentilucci U, Pedone C, Muley-Vilamu M, et al. The pharmacological treatment of chronic comorbidities in COPD: mind the gap! Pulm Pharmacol Ther 2018; 51:48-58. \u003c/li\u003e\n\u003cli\u003eGarvey C, Criner GJ. Impact of comorbidities on the treatment of chronic obstructive pulmonary disease. Am J Med 2018; 131(9S):23-29. \u003c/li\u003e\n\u003cli\u003est\u0026auml;llberg B, Janson C, Larsson K, et al. Real-world retrospective cohort study ARCTIC shows the burden of comorbidities in Swedish COPD versus non-COPD patients. Prim Care Respir Med 2018; 28(1):33. \u003c/li\u003e\n\u003cli\u003egazzotti M, Roco C, Pradella C, et al. Frequency of osteoporosis and vertebral fractures in chronic obstructive pulmonary disease (COPD) patients. Arch Bronconeumol 2019; 55(5):252-257. \u003c/li\u003e\n\u003cli\u003em. Brennan, d. nash, r. rutherford. 87 vertebral fractures in older patients with COPD: an under-diagnosed and under-treated entity age aging. 51. Supple3. 2022. \u003c/li\u003e\n\u003cli\u003eDou Z, Chen X, Chen J, Yang H, Chen J. Chronic Obstructive Pulmonary Disease and Osteoporosis: a Two-Sample Mendelian Randomization Analysis. Chronic Obstr Pulm Dis. 2024 Jul 25;11(4):416-426.\u003c/li\u003e\n\u003cli\u003eShen L, Lv J, Li J, Zhou J, Wang X. Managing Osteoporosis in COPD. Endocr Metab Immune Disord Drug Targets. 2024;24(8):896-901.\u003c/li\u003e\n\u003cli\u003eFogelman l, Blake GM. Different approaches to bone densitometry. JNucl Med 2000; 41(12):2015-2025.\u003c/li\u003e\n\u003cli\u003ePeng T, Zeng X, Li Y, Li M, Pu B, Zhi B, Wang Y, Qu H. A study on whether deep learning models based on CT images for bone density classification and prediction can be used for opportunistic osteoporosis screening. Osteoporos Int. 2024 Jan;35(1):117-128. \u003c/li\u003e\n\u003cli\u003eLambin P Leijenaar RTH, DeistTM et al. (2017) Radiomics: the bridge between medical imaging and personalized medicine. Nat Rev ClinOncol 14(12):749- 762.\u003c/li\u003e\n\u003cli\u003eMayerhoefer ME, Materka A, Langs G et al (2020) Introduction to radi-omics.JNucl Med 61(4):488-495.\u003c/li\u003e\n\u003cli\u003eZhu Z, Zhao S, Li J, Wang Y, Xu L, Jia Y, Li Z, Li W, Chen G, Wu X. Development and application of a deep learning-based comprehensive early diagnostic model for chronic obstructive pulmonary disease. Respir Res. 2024 Apr 18;25(1):167. \u003c/li\u003e\n\u003cli\u003eMakimoto K, Hogg JC, Bourbeau J, Tan WC, Kirby M; CanCOLD Collaborative Research Group. Enhancing COPD classification using combined quantitative computed tomography and texture-based radiomics: a CanCOLD cohort study. ERJ Open Res. 2024 Jul 22;10(4):00968-2023. \u003c/li\u003e\n\u003cli\u003eLin X, Zhou T, Ni J, Li J, Guan Y, Jiang X, Zhou X, Xia Y, Xu F, Hu H, Dong Q, Liu S, Fan L. CT-based whole lung radiomics nomogram: a tool for identifying the risk of cardiovascular disease in patients with chronic obstructive pulmonary disease. Eur Radiol. 2024 Aug;34(8):4852-4863. \u003c/li\u003e\n\u003cli\u003eLin X, Shen R, Zheng X, Shi S, Dai Z, Fang K. Utilizing radiomics techniques to isolate a single vertebral body from chest CT for opportunistic osteoporosis screening. BMC Musculoskelet Disord. 2024 Oct 4;25(1):785.\u003c/li\u003e\n\u003cli\u003eLabaki WW, Rosenberg SR. Chronic Obstructive Pulmonary Disease. Ann Intern Med. 2020 Aug 4;173(3): ITC17-ITC32.\u003c/li\u003e\n\u003cli\u003eHalpin DMG, Criner GJ, Papi A, Singh D, Anzueto A, Martinez FJ, Agusti AA, Vogelmeier CF. Global Initiative for the Diagnosis, Management, and Prevention of Chronic Obstructive Lung Disease. The 2020 GOLD Science Committee Report on COVID-19 and Chronic Obstructive Pulmonary Disease. Am J Respir Crit Care Med. 2021 Jan 1;203(1):24-36. \u003c/li\u003e\n\u003cli\u003eBrown KH, Ghita-Pettigrew M, Kerr BN, Mohamed-Smith L, Walls GM, McGarry CK, Butterworth KT. Characterization of quantitative imaging biomarkers for inflammatory and fibrotic radiation-induced lung injuries using preclinical radiomics. Radiother Oncol. 2024 Mar;192:110106. \u003c/li\u003e\n\u003cli\u003eAdas-Okuma MG, Maeda SS, Gazzotti MR, Roco CM, Pradella CO, Nascimento OA, Porto EF, Vieira JGH, Jardim JR, Lazaretti-Castro M. COPD as an independent risk factor for osteoporosis and fractures. Osteoporos Int. 2020 Apr;31(4):687-697. \u003c/li\u003e\n\u003cli\u003eBitar AN, Sulaiman SAS, Ali IABH, Khan AH. Prevalence, risk assessment, and predictors of osteoporosis among chronic obstructive pulmonary disease patients. J Adv Pharm Technol Res. 2021 Oct-Dec;12(4):395-401. \u003c/li\u003e\n\u003cli\u003eChen YW, Ramsook AH, Coxson HO, et al. Prevalence and risk factors for osteoporosis in individuals with COPD: a systematic review and meta-analysis. Chest. 2019;156(6):1092-1110. \u003c/li\u003e\n\u003cli\u003eBorgstr\u0026ouml;m F, Karlsson L, Orts\u0026auml;ter G, et al. Fragility fractures in Europe: burden, management, and opportunities. Arch Osteoporos. 2020;15(1):59. \u003c/li\u003e\n\u003cli\u003eLi Y, Gao H, Zhao L, Wang J. Osteoporosis in COPD patients: risk factors and pulmonary rehabilitation. Clin Respir J. 2022 Jul;16(7):487-496. \u003c/li\u003e\n\u003cli\u003eZhang B, Chen Z, Yan R, Lai B, Wu G, You J, Wu X, Duan J, Zhang S. Development and Validation of a Feature-Based Broad-Learning System for Opportunistic Osteoporosis Screening Using Lumbar Spine Radiographs. Acad Radiol. 2024 Jan;31(1):84-92. \u003c/li\u003e\n\u003cli\u003eZhen T, Fang J, Hu D, Shen Q, Ruan M. Comparative evaluation of multiparametric lumbar MRI radiomic models for detecting osteoporosis. BMC Musculoskelet Disord. 2024 Feb 29;25(1):185. \u003c/li\u003e\n\u003cli\u003eFang K, Zheng X, Lin X, Dai Z. A comprehensive approach for osteoporosis detection through chest CT analysis and bone turnover markers: harnessing radiomics and deep learning techniques. Front Endocrinol (Lausanne). 2024 Jun 4;15:1296047.\u003c/li\u003e\n\u003cli\u003ePan J, Lin PC, Gong SC, Wang Z, Cao R, Lv Y, Zhang K, Wang L. Feasibility study of opportunistic osteoporosis screening on chest CT using a multi-feature fusion DCNN model. Arch Osteoporos. 2024 Oct 17;19(1):98. \u003c/li\u003e\n\u003cli\u003eYang H, Li Y, Yang H, Shi Z, Yao Q, Jia C, Song M, Qin J. A Novel CT-Based Fracture Risk Prediction Model for COPD Patients. Acad Radiol. 2024 Oct 10:S1076- 6332(24)00602-0. \u003c/li\u003e\n\u003cli\u003eZhu Z, Zhao S, Li J, Wang Y, Xu L, Jia Y, Li Z, Li W, Chen G, Wu X. Development and application of a deep learning-based comprehensive early diagnostic model for chronic obstructive pulmonary disease. Respir Res. 2024 Apr 18;25(1):167.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"chronic obstructive pulmonary disease, osteoporosis, radiomics, predictive modeling, computed tomography","lastPublishedDoi":"10.21203/rs.3.rs-6456138/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6456138/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTo evaluate the value of CT whole-lung imaging histograms in diagnosing osteoporosis (OP) risk in patients with chronic obstructive pulmonary disease (COPD).258 COPD patients were divided into a training cohort (n\u0026thinsp;=\u0026thinsp;149), an internal validation cohort (n\u0026thinsp;=\u0026thinsp;64), and an external validation cohort (n\u0026thinsp;=\u0026thinsp;45). Clinical data and CT results were analyzed. Imaging histologic features of the whole lung were extracted from chest CT images. Machine learning algorithms were utilized to construct the radiomics model. Multifactor logistic regression analysis was used to build the radiomics nomogram by combining independent clinical factors. ROC curves were used to analyze the predictive performance of the models.We developed a model to predict osteoporosis risk in patients with COPD by integrating imaging histologic features, as well as independent clinical risk factors. On the training set, the joint model (area under the curve [AUC], 0.811), the clinical model (AUC, 0.691), and the imaging model (AUC, 0.762). On the internal validation set, the joint model (AUC, 0.806), the clinical model (AUC, 0.724), and the imaging model (AUC, 0.765). On the external validation set, the joint model (AUC, 0.728), the clinical model (AUC, 0.656), and the imaging model (AUC, 0.718). Decision curve analysis showed that the joint model was superior to the single radiomics model with clinical factors. CT-based whole-lung radiomics nomograms are valuable in diagnosing the risk of osteoporosis in patients with COPD.\u003c/p\u003e","manuscriptTitle":"The potential value of CT-based whole lung radiomics nomogram for predicting osteoporosis risk in COPD patients: a two-center study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-16 13:21:00","doi":"10.21203/rs.3.rs-6456138/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a7d607fa-2f4f-4b02-882d-c38abf2a81ed","owner":[],"postedDate":"May 16th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":48494550,"name":"Health sciences/Endocrinology/Endocrine system and metabolic diseases"},{"id":48494551,"name":"Health sciences/Biomarkers"},{"id":48494552,"name":"Health sciences/Endocrinology"},{"id":48494553,"name":"Health sciences/Medical research"},{"id":48494554,"name":"Health sciences/Risk factors"}],"tags":[],"updatedAt":"2026-01-23T06:40:19+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-16 13:21:00","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6456138","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6456138","identity":"rs-6456138","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}