A Novel Clinical Prediction Model for Pulmonary Hypertension Based on Computed Tomography Angiography, Laboratory Data, and basic demographic information

A Novel Clinical Prediction Model for Pulmonary Hypertension Based on Computed Tomography Angiography, Laboratory Data, and basic demographic information | 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 A Novel Clinical Prediction Model for Pulmonary Hypertension Based on Computed Tomography Angiography, Laboratory Data, and basic demographic information Qian Cheng, Xiaojun Hao, Hao Li, Hongxia Jiang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6845563/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 Pulmonary hypertension (PH) is defined as mean pulmonary artery pressure (mPAP) > 20 mmHg and diagnosed invasively via right heart catheterization (RHC). In this study, we developed a noninvasive PH prediction model. Recursive feature elimination (RFE) selected 11 CTA indicators (RESV, LPA, PA, DHDAD, REDV, RSV, RPA, RCO, AAD, REF, and DAD) as predictors. Among ten machine learning models, XGBoost performed best, with SHAP analysis highlighting MPA as the most influential variable. The CTA model achieved high accuracy (training: 97.9%, validation: 90.9%) and robust metrics (AUC > 0.875). Univariate logistic regression identified additional predictors (sex, age, platelet volume, fibrinogen), which, combined with CTA data, improved performance (AUC: training 0.998, validation 0.909). The final logistic regression model with L1 regularization was visualized as a nomogram. Decision curve analysis confirmed clinical utility. This noninvasive approach, integrating CTA, lab tests, and demographics, aids PH diagnosis in RHC-contraindicated or resource-limited settings. Health sciences/Medical research/Experimental models of disease Health sciences/Medical research/Pre clinical studies Pulmonary hypertension machine learning predictive model Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Pulmonary hypertension (PH) is defined as a pathologically elevated mean pulmonary artery pressure (mPAP)>20 mmHg, which is characterized by vascular remodeling, including obstruction, stiffening, and vasoconstriction of the pulmonary vasculature [ 1 ] . The prevalence of PH varies globally, with approximately 1% of the general population affected, and this rate increases to 10% in individuals over the age of 65 years, leading to a higher risk of mortality [ 2 ] . PH is associated with significant morbidity and mortality due to its potential to progress to right heart failure if left untreated [ 3 ] . The gold standard for diagnosing PH is invasive right heart catheterization (RHC), which directly measures hemodynamic parameters such as mPAP and pulmonary vascular resistance (PVR) [ 4 ] . However, RHC is limited by its invasiveness, high cost, and the risk of complications, making it less accessible for widespread clinical use [ 5 ] . Moreover, RHC is not feasible in patients with contraindications or those residing in remote areas where advanced diagnostic facilities are unavailable [ 6 ] . Therefore, there is a pressing need for non-invasive diagnostic tools that can accurately predict PH and guide therapeutic decisions. In recent years, advances in medical imaging and machine learning have opened new avenues for the diagnosis and prediction of PH. Computed tomography angiography (CTA) has emerged as a valuable tool for assessing pulmonary vasculature morphology and detecting early signs of vascular remodeling [ 7 ] . A lot of information can be gained from CTA, including pulmonary artery diameter, aorta diameter, the artial and ventricular size and function, these infomation reveals the structure and function characteristics of the heart and pulmonary vasculature. Machine learning algorithms, such as gradient boosting decision trees (GBDTs) and logistic regression models, have demonstrated promising results in predicting PH by analyzing CTA-derived parameters and clinical data [ 8 ] . These approaches provide a potential alternative to invasive RHC, offering a safer and more accessible method for screening and monitoring PH patients. This study aims to develop a novel clinical prediction model for PH combining CTA, laboratory data, as well as basic demographic information such as age and sex/gender, essentially encompassing all the clinically available data with relatively high completeness that we could collect in this study. By leveraging machine learning techniques, we seek to identify key predictors of PH and create a robust model that can accurately diagnose PH. The integration of CTA-derived morphological features with clinical biomarkers represents a significant advancement in the field of PH research, with the potential to improve patient outcomes and reduce the reliance on invasive diagnostic procedures. Data Collection and Statistical Analysis Data Collection From February 2022 to April 2025, we retrospectively collected clinical and imaging data from patients who underwent pulmonary computed tomography angiography (CTA) at our institution. The study cohort consisted of two groups: Pulmonary Hypertension (PH) Group: Patients with confirmed PH (mean pulmonary arterial pressure > 20 mmHg) via right heart catheterization. Control Group: Patients with normal pulmonary CTA findings (no evidence of PH). Collected Variables Demographics (sex, age). CTA Parameters(Cardiac Morphology: Right atrial enlargement, right ventricular (RV) enlargement, left atrial (LA) enlargement, left ventricular (LV) enlargement, atrial septal defect, ventricular septal defect. Ventricular Dimensions: RV transverse diameter, RV wall thickness, RA superior-inferior diameter, RA left-right diameter, LA anteroposterior diameter, LA superior-inferior diameter, LA left-right diameter, LV transverse diameter.Functional Metrics: LV end-diastolic volume (LEDV), LV end-systolic volume (LESV), LV stroke volume (LSV), LV ejection fraction (LEF), LV mass (Leftmass), LV cardiac output (LeftCO), RV end-diastolic volume (REDV), RV end-systolic volume (RESV), RV stroke volume (RSV), RV ejection fraction (REF), RV cardiac output (Right CO). Vascular Measurements: Pulmonary artery (PA) diameter, left PA diameter (LPA), right PA diameter (RPA), McGoon ratio, Aortic dimensions: Sinotubular junction diameter, ascending aorta diameter, descending aorta diameter, diaphragmatic-level descending aorta diameter ,Total calcium score. ) Laboratory Parameters(Complete Blood Count: White blood cells, red blood cells, hemoglobin, platelets, neutrophil percentage, lymphocyte percentage, monocyte percentage, eosinophil percentage, basophil percentage, absolute neutrophil count, absolute lymphocyte count, absolute monocyte count, absolute eosinophil count, absolute basophil count, hematocrit, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), red cell distribution width (RDW-CV, RDW-SD), mean platelet volume (MPV). Coagulation Profile: Prothrombin time (PT), international normalized ratio (INR), prothrombin activity, activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen. D-dimer. NT-proBNP. Procalcitonin (PCT). Liver and Renal Function + Electrolytes: Alanine aminotransferase (ALT), aspartate aminotransferase (AST), ALT/AST ratio, direct bilirubin, indirect bilirubin, total bilirubin, total protein, albumin, globulin, albumin-to-globulin ratio (A/G), gamma-glutamyl transferase (GGT), alkaline phosphatase (ALP), total bile acids, creatinine, blood urea nitrogen (BUN), uric acid, CO₂, cystatin C, potassium, sodium, chloride, calcium, magnesium, phosphorus. Lipid Profile: Total cholesterol, triglycerides, high-density lipoprotein (HDL), low-density lipoprotein (LDL), small dense LDL, lipoprotein(a), free fatty acids, phospholipids. Thyroid Function: Free T3, free T4, thyroid-stimulating hormone (TSH). Immunological Markers: Anti-cardiolipin antibodies (IgM, IgG, IgA), complement C3, C4, immunoglobulins (IgG, IgA, IgM, IgE), anti-streptolysin O (ASO), rheumatoid factor (RF), RF antibodies (IgM, IgG, IgA). Arterial Blood Gas Analysis: pH, partial pressure of oxygen (PaO₂), oxygen saturation (SaO₂), partial pressure of CO₂ (PaCO₂), temperature, lactate, standard bicarbonate, actual base excess, standard base excess, anion gap. Statistical Analysis Data Preprocessing Variables or samples with > 30% missing data were excluded. Remaining missing values were imputed using the k-nearest neighbors (KNN) method. The final dataset included 59 PH patients and 43 controls, with 84 variables retained for analysis. Descriptive Statistics Continuous variables were assessed for normality using the Shapiro-Wilk test (p > 0.05 considered normally distributed). Normally distributed data: Expressed as mean ± standard deviation (SD); compared using independent t-tests (Levene’s test for homogeneity of variance, p > 0.05) or Welch’s t’-test (if variances unequal). Non-normally distributed data: Expressed as median (Q1, Q3); compared using the Mann-Whitney U test. Categorical variables: Binary variables: Compared using Pearson’s chi-square test. Multiclass variables: Compared using the Kruskal-Wallis test. Statistical significance was set at p < 0.05. All analyses were performed using IBM SPSS Statistics 23. Machine Learning Modeling Data Splitting: Stratified sampling divided the dataset into training (80%) and testing (20%) sets. Feature Selection: Recursive feature elimination (RFE) with 10-fold cross-validation was applied using a random forest-based approach (R caret package). Model Training: Multiple machine learning algorithms were evaluated: LASSO, elastic net, decision tree, random forest, XGBoost, SVM, KNN, naive Bayes, gradient boosting machine (GBM). Model performance was assessed via area under the ROC curve (AUC). Model Interpretability: SHapley Additive exPlanations (SHAP) analysis (Python SHAP package) identified key predictive features. Validation: Bootstrap resampling (1,000 iterations) evaluated model stability. Performance metrics included: Accuracy, AUC, F1-score, precision, recall, specificity. Final Predictive Model Univariate logistic regression identified significant predictors from clinical/lab variables. These predictors were combined with the optimal CTA-derived features to construct an L1-regularized (LASSO) logistic regression model. A nomogram was developed (R RMS package) to visualize individual risk contributions. Decision curve analysis (DCA) (R RMDA package) assessed clinical utility. Calibration curves and AUC were computed to evaluate model performance in training and test sets. Results 1. Baseline Characteristics of the Study Population The overall data analysis workflow of this study has been described in detail (Fig. 1 ). A total of 102 patients were ultimately enrolled in this study, including 59 patients (57.8%) with pulmonary hypertension (PH) confirmed by right heart catheterization and 43 controls (42.2%). A comparison of baseline characteristics between the two groups (Table 1 ) revealed that PH patients had a lower proportion of males (30.5% vs. 60.5%, p = 0.003). Regarding imaging parameters, the PH group exhibited typical PH characteristics: the main pulmonary artery diameter was significantly increased (36.64 ± 8.7 mm vs. 26.51 ± 7 mm, p < 0.001), right ventricular end-diastolic volume (RVEDV) was markedly enlarged (229.06 ± 86.57 mL vs. 173.52 ± 63.87 mL, p = 0.03), and right ventricular end-systolic volume (RVESV) was significantly higher (123.48 [96.2–169.88] mL vs. 79.1 [59.38–88.85] mL, p = 0.001). Notably, all cases of right atrial enlargement (43 cases) and right ventricular enlargement (41 cases) were observed exclusively in the PH group (p < 0.001), consistent with the hemodynamic burden imposed by PH. Table 1 Baseline Characteristics of the Variables Variable Name Total (n = 102) Control Group (n = 43) Case Group (n = 59) p-value Age, years 51.46 (33.55-68) 66.74 (51.46–75.5) 42 (30-56.5) <0.001 Male gender (%) 44(43.1%) 26(60.5%) 18(30.5%) 0.003 Right atrial enlargement, n (%) 43 0 43 <0.001 Right ventricular enlargement, n (%) 41 0 41 <0.001 Left atrial enlargement, n (%) 8 0 8 0.012 Left ventricular enlargement, n (%) 4 0 4 0.082 Atrial septal defect, n (%) 11 0 11 0.003 Ventricular septal defect, n (%) 7 0 7 0.019 Left ventricular end-diastolic volume (LVEDV), ml 120.07 (97.3-151.03) 115.29 (94.92-131.96) 121.62 (99.94-162.61) 0.176 Left ventricular end-systolic volume (LVESV), ml 38.25 (27.64–50.69) 33.66 (24.27–44.34) 41.56 (30.95–63.72) 0.029 Left ventricular stroke volume (LVSV), ml 81.36 (66.85–97.28) 81.54 (67.91–92.89) 81.18 (67.1-104.94) 0.595 Left ventricular ejection fraction (LVEF), % 70 (61–76) 73 (64.5–76) 66 (59.5–75) 0.053 Left ventricular cardiac output, L/min 5.73 (4.57–7.58) 5.84 (4.56–7.54) 5.71 (4.62–8.09) 0.721 Right ventricular end-diastolic volume (RVEDV), ml 218.75 ± 85.24 173.52 ± 63.87 229.06 ± 86.57 0.03 Right ventricular end-systolic volume (RVESV), ml 112.76 (79.83-159.65) 79.1 (59.38–88.85) 123.48 (96.2-169.88) 0.001 Right ventricular stroke volume (RVSV), ml 89.75 (67.8-124.04) 72.55 (61.85-104.83) 90.23 (70.71-124.18) 0.282 Right ventricular ejection fraction (RVEF), % 45.87 ± 16.24 47.14 ± 17.68 45.57 ± 16.05 0.76 Right ventricular cardiac output, L/min 6.28 (4.2–8.89) 4.49 (3.91–7.08) 6.33 (4.64–8.95) 0.162 Pulmonary artery diameter, mm 34.45 ± 9.32 26.51 ± 7 36.64 ± 8.7 <0.001 Left pulmonary artery diameter, mm 20.42 ± 5.77 16.58 ± 3.75 21.35 ± 5.81 0.001 Right pulmonary artery diameter, mm 22.14 ± 7.25 16.86 ± 5.22 23.42 ± 7.12 0.001 Ascending aorta diameter, mm 30 (24–33) 32 (30–34) 25 (22.5–30.5) <0.001 Descending aorta diameter, mm 18 (16–21) 19.85 (18.25-20) 18 (16–21) 0.487 Diaphragmatic level descending aorta diameter, mm 16.88 ± 3.48 17.18 ± 4.49 16.81 ± 3.23 0.73 White blood cell count, 10⁹/L 5.64 (4.86–7.08) 5.5 (4.93–7.3) 5.91 (4.8–6.91) 0.704 Red blood cell count, 10⁹/L 4.26 (3.79–4.8) 4.08 (3.58–4.44) 4.35 (3.98–4.94) 0.005 Hemoglobin concentration, g/L 127 (110–144) 126 (107.05-139.75) 133 (113.5-150.5) 0.087 Platelet count, 10⁹/L 191.5 (135.5-238.75) 213 (173.5–268) 175 (117-216.5) 0.006 Neutrophil percentage, % 62.5 ± 11.29 64.4 ± 14.02 61.11 ± 8.67 0.15 Lymphocyte percentage, % 27.02 ± 10.39 25.08 ± 12.48 28.43 ± 8.38 0.11 Monocyte percentage, % 7.8 (6.2–9.3) 7.8 (7-9.45) 7.8 (5.9–9.2) 0.382 Eosinophil percentage, % 1.3 (0.7–2.5) 1.3 (0.7–2.6) 1.2 (0.7–2.3) 0.504 Basophil percentage, % 0.5 (0.3–0.8) 0.6 (0.35–0.8) 0.5 (0.3–0.8) 0.65 Absolute neutrophil count, 10⁹/L 3.62 (2.79–4.6) 3.7 (2.81–5.37) 3.54 (2.72–4.43) 0.173 Absolute lymphocyte count, 10⁹/L 1.5 (1.12–2.06) 1.37 (1–2) 1.6 (1.22–2.08) 0.099 Absolute monocyte count, 10⁹/L 0.44 (0.37–0.6) 0.46 (0.4–0.72) 0.43 (0.33–0.54) 0.147 Absolute eosinophil count, 10⁹/L 0.1 (0.03–0.14) 0.1 (0.04–0.2) 0.1 (0.02–0.11) 0.186 Absolute basophil count, 10⁹/L 47 (46.08) 23 (53.49) 24 (40.68) 0.288 Hematocrit, % 38.65 (33.73–43.15) 37.3 (32.5–41.3) 40.6 (35.35–44.15) 0.02 Mean corpuscular volume (MCV), fL 91.3 (87.3–95.3) 91.5 (88.7-95.85) 90.65 (85.95–94.65) 0.266 Mean corpuscular hemoglobin (MCH), pg 30.3 (28.13–31.67) 30.5 (28.55–31.85) 30 (28.1-31.55) 0.272 Mean corpuscular hemoglobin concentration (MCHC), g/L 330.5 (321.05–337) 332 (323.25–337) 330 (317.5-336.5) 0.139 Red cell distribution width-CV (RDW-CV), % 14 (13.2-15.75) 13.5 (13.2-15.15) 14.35 (13.38–16.15) 0.087 Mean platelet volume (MPV), fL 9 (8.1–10.5) 8.5 (7.95–9.3) 9.45 (8.38–11.1) 0.001 Prothrombin time (PT), seconds 12.1 (11.4–13.1) 11.9 (11.15–13.3) 12.2 (11.6–13) 0.255 International Normalized Ratio (INR) 1.11 (1.05–1.21) 1.09 (1.02–1.23) 1.12 (1.06–1.19) 0.314 Prothrombin activity, % 84 (75–94) 87 (73-96.5) 84 (77-91.75) 0.496 Activated partial thromboplastin time (APTT), seconds 32.2 (29.9–34.4) 32.1 (29.65–34.45) 32.3 (30.18–34.35) 0.789 Thrombin time (TT), seconds 14.77 ± 1.36 14.62 ± 1.34 14.88 ± 1.38 0.34 Fibrinogen concentration, mg/dL 284 (252.5–330) 300 (267–350) 275.5 (242-310.5) 0.015 D-dimer, ng/mL 214.5 (101–760) 504 (160.5–1333) 153 (85–260) <0.001 N-terminal pro-brain natriuretic peptide (NT-proBNP), U/L 279 (88.45–872.5) 195 (72.3–569) 375 (103.25–994) 0.136 Alanine aminotransferase (ALT), U/L 20 (13–28) 20 (12–28) 20 (14–28) 0.69 Aspartate aminotransferase (AST), U/L 22 (18–29) 22 (17–25) 23 (19–31) 0.215 AST/ALT ratio 1.29 (0.95–1.6) 1.27 (0.86–1.77) 1.33 (1-1.54) 0.844 Total bilirubin, µmol/L 13.8 (9.4-18.67) 12.1 (8.65–15.95) 15.2 (10.3–20.9) 0.022 Direct bilirubin, µmol/L 3.7 (2.7–5.2) 3.4 (2.25–4.4) 3.9 (2.95–6.85) 0.138 Indirect bilirubin, µmol/L 9.8 (6.2-13.15) 8.5 (6.05–10.85) 11.1 (7.1–14.6) 0.013 Total protein, g/L 67.6 (63.7–72.9) 66.8 (61.87–70.65) 68.8 (64.45–73.2) 0.055 Albumin, g/L 39.2 (36.53–41.97) 37.65 (34.6-40.98) 39.95 (37.42–42.75) 0.012 Globulin, g/L 28.3 (25.3–32.3) 28 (25.25–32.85) 28.65 (25.38–31.85) 0.997 Albumin/Globulin ratio (A/G ratio) 1.33 (1.18–1.58) 1.29 (1.13–1.52) 1.34 (1.23–1.6) 0.194 Gamma-glutamyl transferase (GGT), U/L 27 (16–58) 22 (13.5–40) 32 (20–60) 0.083 Alkaline phosphatase (ALP), U/L 74 (60.25-95) 80 (60.5–104) 72 (60.5–88) 0.222 Total bile acid (TBA), µmol/L 4.75 (2.8–9.07) 3.7 (2-7.12) 5 (2.95–12.1) 0.051 Blood urea nitrogen (BUN), mmol/L 5.52 (4.4–6.97) 5.7 (4.28–8.03) 5.3 (4.43–6.5) 0.492 Serum creatinine, µmol/L 67.65 (57.05–82.75) 74 (60.65–89.65) 64.7 (53.65–80.55) 0.046 Uric acid, µmol/L 343.5 (257-413.7) 309 (243.55–384.2) 362.35 (276.62–483) 0.071 Carbon dioxide (CO₂), mmol/L 22.9 (20.7–25.7) 23.6 (21.25–26.15) 22.8 (20.4-25.05) 0.29 Serum cystatin C, mg/L 0.99 (0.78–1.17) 1.01 (0.83–1.23) 0.91 (0.77–1.08) 0.117 Potassium, mmol/L 3.98 ± 0.44 4.03 ± 0.48 3.94 ± 0.41 0.33 Sodium, mmol/L 139.7 (137.4-141.1) 139.4 (137.4-141.9) 139.85 (137.55-140.78) 0.678 Chloride, mmol/L 105.2 (102.75-107.65) 104.6 (101.55-107.45) 105.3 (103-107.75) 0.245 Calcium, mmol/L 2.27 (2.19–2.35) 2.23 (2.16–2.31) 2.3 (2.21–2.39) 0.002 Magnesium, mmol/L 0.84 (0.78–0.91) 0.86 (0.8–0.92) 0.82 (0.78–0.9) 0.218 Phosphate, mmol/L 1.19 (1.05–1.3) 1.13 (1-1.26) 1.22 (1.1–1.33) 0.03 Total cholesterol, mmol/L 4.15 ± 0.93 3.9 ± 0.85 4.31 ± 0.96 0.05 Triglycerides, mmol/L 1.35 (0.95–2.07) 1.07 (0.92–1.77) 1.44 (1-2.35) 0.079 High-density lipoprotein cholesterol (HDL-C), mmol/L 1.04 (0.88–1.22) 0.99 (0.84–1.19) 1.06 (0.9–1.3) 0.296 Low-density lipoprotein cholesterol (LDL-C), mmol/L 2.47 ± 0.77 2.4 ± 0.69 2.51 ± 0.82 0.55 Small dense LDL cholesterol (sdLDL-C), mmol/L 0.68 (0.48–1.18) 0.58 (0.46–0.81) 0.88 (0.5–1.26) 0.068 Lipoprotein(a), mmol/L 89.2 (27-180.05) 70.45 (16.48-142.07) 113.5 (39.75-208.15) 0.137 Free fatty acids (FFA), µmol/L 393.4 (270.76–605.7) 437.5 (360.3-562.3) 334.08 (181.2-611.22) 0.153 Phospholipids, mmol/L 2.16 (1.89–2.53) 2.02 (1.85–2.16) 2.36 (1.99–2.7) 0.002 Laboratory test results revealed distinct biochemical characteristics in PH patients: mean platelet volume (MPV) was significantly increased (9.45 [8.38–11.1] fL vs. 8.5 [7.95–9.3] fL, p = 0.001), fibrinogen levels were decreased (275.5 [242–310.5] mg/dL vs. 300 [267–350] mg/dL, p = 0.015), and D-dimer levels were markedly elevated (504 [160.5–1333] ng/mL vs. 153 [85–260] ng/mL, p < 0.001), suggesting a significant coagulation dysfunction and prothrombotic state in PH patients. Additionally, albumin levels (39.95 [37.42–42.75] g/L vs. 37.65 [34.6–40.98] g/L, p = 0.012) and total bilirubin levels (15.2 [10.3–20.9] µmol/L vs. 12.1 [8.65–15.95] µmol/L, p = 0.022) demonstrated abnormal changes in the PH group, reflecting the pathophysiological alterations of hepatic congestion and systemic inflammation. These findings provided a solid biological basis for the subsequent development of the prediction model. 2. CTA Feature Selection and Model Optimization Using recursive feature elimination (RFE), 11 optimal predictive features were selected from an initial set of 58 CTA parameters. Ten-fold cross-validation confirmed that model performance peaked when using 11 features, achieving an AUC of 0.89 (Fig. 2 A). As the number of features increased from 5 to 11, the model’s performance improved rapidly (AUC from 0.72 to 0.89). However, including more than 11 features led to a decline in AUC to 0.87, indicating a risk of overfitting. This result suggests that these 11 CTA parameters sufficiently captured the imaging characteristics of PH. The importance ranking of these 11 key CTA parameters was as follows (Fig. 2 B): right ventricular end-systolic volume (RVESV, Gini importance = 0.32), left pulmonary artery diameter (LPA, 0.28), main pulmonary artery diameter (PA, 0.25), diaphragmatic-level descending aorta diameter (DHDAD, 0.22), right ventricular end-diastolic volume (RVEDV, 0.19), right ventricular stroke volume (RVSV, 0.18), right pulmonary artery diameter (RPA, 0.17), right cardiac output (RCO, 0.15), ascending aorta diameter (AAD, 0.13), right ventricular ejection fraction (RVEF, 0.11), and descending aorta diameter (DAD, 0.09). Among ten machine learning algorithms compared (Fig. 2 C), XGBoost demonstrated superior predictive performance (training set AUC = 0.92, sensitivity = 86%, specificity = 89%), significantly outperforming traditional statistical methods such as LASSO regression (AUC = 0.85) and other tree-based models like random forests (AUC = 0.88). The advantage of XGBoost likely stems from its gradient boosting mechanism and regularization strategies, which effectively capture complex nonlinear relationships between CTA parameters and PH. 3. Interpretability Analysis of the CTA Model A beeswarm plot visualized the impact of each feature on model output through SHAP values (Fig. 3 A). The main pulmonary artery (MPA) demonstrated the widest SHAP value distribution with the largest absolute values (approaching ± 1), confirming its role as the strongest predictive factor. The left pulmonary artery (LPA) showed a distinctive negative effect (primarily distributed in the negative SHAP range). Right heart function parameters (RCO, RVESV, RVEDV) were concentrated in the positive range, consistent with clinical pathophysiology. The density of data points revealed a pronounced rightward shift when MPA exceeded 35 mm, supporting this threshold’s clinical relevance. A feature importance bar chart based on mean SHAP values (Fig. 3 B) further confirmed this hierarchy: MPA ranked first with a value of 0.92, followed by LPA (0.91, but negative impact), right ventricular systolic parameters (RVSV 0.72, RVESV 0.65) significantly exceeded diastolic parameters (RVEDV 0.51), while RPA (0.28) and DHDAD (0.22) were of secondary importance. This ranking was corroborated by the beeswarm plot, together establishing a robust hierarchy of feature importance. A heatmap of individual feature contributions across 80 cases (Fig. 3 C) demonstrated inter-individual heterogeneity. The MPA (top row) consistently contributed substantially (deep color blocks), LPA (second row) displayed significant negative contributions in approximately one-third of patients (light color blocks), and right heart parameters (RVSV, RVESV, RCO) exhibited diverse contribution patterns, reflecting phenotypic heterogeneity. A single-sample decision plot (Fig. 3 D) illustrated the prediction pathway for a specific patient. With a baseline E[f(x)] = 1.077 and a predicted f(x) = 3.12 (high risk), MPA (+ 1.149) and RCO (+ 3.498) were the major positive contributors, while LPA (–0.842) exerted a significant negative effect. Other parameters contributed in accordance with overall patterns (RVESV + 0.716, RPA + 1.8), intuitively illustrating the multi-parameter decision logic of the clinical prediction model. 4. CTA Model Validation and Performance Evaluation The CTA model exhibited excellent predictive performance in both the training and validation cohorts. In the training cohort, the accuracy for predicting PH was 97.9%, and 97.1% for controls (Fig. 4 A). In the validation cohort, the accuracy was 90.9% for PH prediction and 87.5% for controls (Fig. 4 B). In the training set (n = 82), all evaluation metrics exceeded 0.97: accuracy 97.9%, AUC 0.998 (95% CI: 0.996–1.000), F1 score 0.978, sensitivity 0.971, and specificity 0.987 (Fig. 4 C). In the independent validation set (n = 20), the model maintained good generalizability: accuracy 90.9%, AUC 0.909 (95% CI: 0.832–0.986), F1 score 0.875, sensitivity 0.875, and specificity 0.917 (Fig. 4 D). ROC curves, precision-recall curves, specificity-sensitivity plots, and accuracy-threshold curves further demonstrated the model’s robust performance in both cohorts (Figs. 4 E– 4 L). 5. Multimodal Model Integration and Optimization Univariate logistic regression identified four independent laboratory predictors: sex, age, MPV, and fibrinogen level. Integrating these factors with the CTA model produced a multimodal prediction model with superior clinical utility (Fig. 5 A). Decision curve analysis (DCA) indicated that laboratory model (Fig. 5 B) yielded positive net clinical benefits across a 10–90% threshold probability range, the integrated model consistently outperformed laboratory model (Fig. 5 C). Notably, within the clinically ambiguous 30–70% decision threshold, the integrated model identified an additional 15–18 true PH patients per 100 cases while reducing 10–12 unnecessary right heart catheterizations, demonstrating substantial clinical value. Bootstrap resampling confirmed the excellent performance of the integrated model (Figs. 5 D– 5 E). 6. Nomogram Development and Clinical Application A nomogram was developed based on L1-regularized logistic regression incorporating five core predictors (Fig. 6 A). Point allocations and total score ranges were as follows: sex: female, 13 points; male, 0 points. Age: total 0–93 points (0–100 years old). MPV: total 0–70 points (6.5–13 fL). Fibrinogen: total 0–100 points (150–650 mg/dL). CTA index score: total 0–98 points (range 0–1). The overall score range was 0 (all optimal values) to 374 (all worst values). Risk stratification was defined as low risk (0–125, PH probability 70%). Calibration curves showed good agreement in both training (Fig. 6 B) and validation cohorts (Fig. 6 C), with an overall mean absolute error of only 0.048, indicating excellent calibration. The model exhibited outstanding discriminatory ability in the training set (AUC = 0.998) (Fig. 6 D) and remained stable in the validation set (AUC = 0.909) (Fig. 6 E). A clinical application example: a 72-year-old female with an MPV of 12.8 fL, fibrinogen of 180 mg/dL, and a CTA index score of 0.82, achieved a total score of 309 (13 [sex] + 67 [age] + 63 [MPV] + 86 [fibrinogen] + 80 [CTA index] = 309). This score (> 250) classified her as high risk with a PH probability exceeding 90%. Clinical recommendations included immediate right heart catheterization and initiation of targeted therapy. This tool is especially valuable for PH screening and referral decision-making in primary healthcare settings. Discussion Constructing of a novel clinical prediction model for the diagnosis of PH based on CTA, laboratory data, and basic demographic information represents a significant breakthrough in the diagnosis and management of this complex condition. Our study leverages advanced machine learning techniques to identify optimal predictors of PH from a comprehensive dataset encompassing both imaging and clinical variables, which nearly encompasses all clinically collectible valuable information.The integration of these diverse data sources enables a more holistic understanding of PH pathophysiology and enhances the accuracy of predictive modeling. Our findings indicate that CTA-derived parameters such as RESV, LPA, PA, DHDAD, REDV, RSV, RPA, RCO, AAD, REF, and DAD are critical predictors of PH. These morphological features reflect the extent of pulmonary vascular remodeling, which is a hallmark of PH pathophysiology [ 9 ] . Selecting XGBOOST as the optimal machine learning model underscores the importance of ensemble methods in capturing complex interactions between variables [ 10 ] . Furthermore, the inclusion of clinical variables such as sex, age, mean platelet volume, and fibrogen content in the final model highlights the multifactorial nature of PH and the need for a comprehensive approach for risk assessment [ 11 – 13 ] . The performance metrics of our model, including accuracy, AUC, F1 score, precision, recall, and specificity, demonstrate exceptional predictive capabilities. Specifically, the model achieved an AUC of 0.998 in the training set and 0.909 in the validation set, indicating high discriminatory power [ 14 ] . This level of performance is comparable to or even superior to existing models that rely solely on clinical or imaging data [ 15 – 17 ] . Moreover, the calibration curve analysis confirms that our model maintains its predictive validity across different patient populations, ensuring reliability in clinical practice . One of the key strengths of our study lies in the rigorous methodology employed to select and validate predictors. Recursive feature elimination (RFE) based on random forests was used to identify the most relevant features from the vast CTA dataset, followed by cross-validation and SHAP analysis to ensure robustness and interpretability [ 18 , 19 ] . This systematic approach minimizes overfitting and ensures that the final model generalizes well to new data. Another highlight of our study is that all patients in the PH groups were confirmed to have pulmonary hypertension via RHC (the gold standard for PH diagnosis), while all control subjects showed no significant structural or functional abnormalities in the cardiovascular system based on CTA findings. The zero misdiagnosis rate in both the case and control groups further enhances the credibility of the final model. The practical implications of our model are substantial. For patients who are contraindicated for RHC or reside in remote areas without access to advanced diagnostic facilities, our model provides a reliable alternative for PH screening. Additionally, the model can inform therapeutic decisions by identifying high-risk patients who require urgent intervention or those who may benefit from targeted treatments such as endothelin receptor antagonists or phosphodiesterase inhibitors . However, several limitations must be acknowledged. The retrospective nature of our study design introduces potential biases related to data collection and patient selection. Future prospective studies are necessary to validate our findings in diverse populations and settings. Furthermore, while our model demonstrates high accuracy, it may not capture all nuances of PH pathophysiology, particularly in cases where comorbidities or genetic factors play a significant role [ 20 , 21 ] . In conclusion, our study presents a novel clinical prediction model for PH that integrates CTA-derived morphological features with laboratory data and basic demographic information. This model represents a significant advancement in non-invasive PH diagnosis and has the potential to improve patient outcomes by facilitating early detection and personalized treatment strategies. Future research should focus on validating this model in larger, more diverse populations and exploring its applicability to other cardiovascular conditions. Abbreviations AUC Area under the receiver operating characteristic curve ALT Alanine aminotransferase APTT Activated partial thromboplastin time AST Aspartate aminotransferase CTA Computed tomography angiography DCA Decision curve analysis HDL High-density lipoprotein INR International normalized ratio LDL Low-density lipoprotein LPA Left pulmonary artery LVEDV Left ventricular end-diastolic volume LVEF Left ventricular ejection fraction LVESV Left ventricular end-systolic volume MPA Main pulmonary artery MPV Mean platelet volume NT-proBNP N-terminal pro-brain natriuretic peptide PA Pulmonary artery PH Pulmonary hypertension PT Prothrombin time RCO Right cardiac output REDV Right ventricular end-diastolic volume REF Right ventricular ejection fraction RESV Right ventricular end-systolic volume RFE Recursive feature elimination RHC Right heart catheterization RPA Right pulmonary artery RSV Right ventricular stroke volume RV Right ventricle/ventricular RVEF Right ventricular ejection fraction SHAP SHapley Additive exPlanations TT Thrombin time. Declarations Ethics statement This study was conducted in accordance with the ethical principles of the Declaration of Helsinki and was approved by the Medical Ethics Committee of Zhongnan Hospital, Wuhan University. All data were anonymized to protect patient confidentiality, and only de-identified information was used for analysis. Due to the retrospective nature of the study, the need to obtain the informed consent was waived by the Medical Ethics Committee of Zhongnan Hospital, Wuhan University. Competing Interests All the authors declare that no conflict of interest exists. Funding This study was funded by the following programs: The National Natural Science Foundation of China (No.82300078), The Fundamental Research Funds for the Central Universities (No.2042023kf0059), Wuhan University Clinical Medicine + Youth Supporting Program (No.413000557). Author Contribution All authors contributed to this article: Qian Cheng contributed to Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft. Xiaojun Hao contributed to Data curation, Formal analysis, Validation, Investigation. Hao Li contributed to Formal analysis, Validation, Investigation, Visualization, Methodology. Hongxia Jiang contributed to Resources, Supervision, Funding acquisition, Project administration, Writing—review & editing, and verifing the underlying data reported in the manuscript. All authors have read and agreed to the published version of the article. Data Availability All data are available from the corresponding author upon reasonable request References Simonneau, G. et al. Haemodynamic definitions and updated clinical classification of pulmonary hypertension. Eur. Respir. J. 53 (1), 1801913 (2019). Galiè, N. et al. 2015 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension . Eur. Respir. J. 46 (4): pp. 903–975 . Kularatne, M. et al. Updated Clinical Classification and Hemodynamic Definitions of Pulmonary Hypertension and Its Clinical Implications . J. Cardiovasc. Dev. Dis. , 11 (3). (2024). Gonzalez-Hermosillo, L. M. et al. Right Heart Catheterization (RHC): A Comprehensive Review of Provocation Tests and Hepatic Hemodynamics in Patients With Pulmonary Hypertension (PH). Curr. Probl. Cardiol. 47 (12), 101351 (2022). Wang, M. T. et al. Application of homocysteine as a non-invasive and effort-free measurements for risk assessment of patients with pulmonary hypertension. Cardiol. J. 31 (2), 285–299 (2024). Singh, V., Dwivedi, S. N. & Deo, S. V. S. Ordinal logistic regression model describing factors associated with extent of nodal involvement in oral cancer patients and its prospective validation. BMC Med. Res. Methodol. 20 (1), 95 (2020). Fisher, M. R. et al. Accuracy of Doppler echocardiography in the hemodynamic assessment of pulmonary hypertension. Am. J. Respir Crit. Care Med. 179 (7), 615–621 (2009). Van den Eynde, J. et al. Successfully implemented artificial intelligence and machine learning applications in cardiology: State-of-the-art review. Trends Cardiovasc. Med. 33 (5), 265–271 (2023). Spruijt, O. A. et al. Predicting pulmonary hypertension with standard computed tomography pulmonary angiography. Int. J. Cardiovasc. Imaging . 31 (4), 871–879 (2015). Zhao, Q. Y. et al. Development and Validation of a Machine-Learning Model for Prediction of Extubation Failure in Intensive Care Units. Front. Med. (Lausanne) . 8 , 676343 (2021). Noris, P., Melazzini, F. & Balduini, C. L. New roles for mean platelet volume measurement in the clinical practice? Platelets 27 (7), 607–612 (2016). Sokou, R. et al. Diagnostic and Prognostic Value of Hematological Parameters in Necrotizing Enterocolitis: A Systematic Review . J. Clin. Med. , 14 (7). (2025). Upadhyay, R. K. Emerging risk biomarkers in cardiovascular diseases and disorders. J Lipids, 2015: p. 971453. (2015). Zhang, N. et al. Machine Learning Based on Computed Tomography Pulmonary Angiography in Evaluating Pulmonary Artery Pressure in Patients with Pulmonary Hypertension . J. Clin. Med. , 12 (4). (2023). Zeng, Y. et al. A novel clinical prediction scoring system of high-altitude pulmonary hypertension. Front. Cardiovasc. Med. 10 , 1290895 (2023). Lui, J. K. et al. A Multimodal Prediction Model for Diagnosing Pulmonary Hypertension in Systemic Sclerosis. Arthritis Care Res. (Hoboken) . 75 (7), 1462–1468 (2023). Nathan, S. D. et al. Derivation and validation of a noninvasive prediction tool to identify pulmonary hypertension in patients with IPF: Evolution of the model FORD. J. Heart Lung Transpl. 43 (4), 547–553 (2024). Ende-Verhaar, Y. M. et al. Usefulness of standard computed tomography pulmonary angiography performed for acute pulmonary embolism for identification of chronic thromboembolic pulmonary hypertension: results of the InShape III study. J. Heart Lung Transpl. 38 (7), 731–738 (2019). Shahin, Y. et al. Quantitative CT Evaluation of Small Pulmonary Vessels Has Functional and Prognostic Value in Pulmonary Hypertension. Radiology 305 (2), 431–440 (2022). Arvanitaki, A. et al. Noninvasive diagnostic modalities and prediction models for detecting pulmonary hypertension associated with interstitial lung disease: a narrative review . Eur. Respir Rev. , 33 (174). (2024). Palazzini, M. et al. Pulmonary hypertension due to left heart disease: analysis of survival according to the haemodynamic classification of the 2015 ESC/ERS guidelines and insights for future changes. Eur. J. Heart Fail. 20 (2), 248–255 (2018). Additional Declarations No competing interests reported. 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-6845563","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":477544814,"identity":"da7a22c1-aa0f-4235-99db-d5cd5be1acf9","order_by":0,"name":"Qian Cheng","email":"","orcid":"","institution":"Zhongnan Hospital of Wuhan University","correspondingAuthor":false,"prefix":"","firstName":"Qian","middleName":"","lastName":"Cheng","suffix":""},{"id":477544815,"identity":"36cfa0fa-ea0a-4af6-97aa-32983382b03e","order_by":1,"name":"Xiaojun Hao","email":"","orcid":"","institution":"Zhongnan Hospital of Wuhan University","correspondingAuthor":false,"prefix":"","firstName":"Xiaojun","middleName":"","lastName":"Hao","suffix":""},{"id":477544816,"identity":"06be8067-a279-4671-afb7-9f4e2abc525e","order_by":2,"name":"Hao Li","email":"","orcid":"","institution":"Zhongnan Hospital of Wuhan University","correspondingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Li","suffix":""},{"id":477544817,"identity":"663d90ce-0716-4914-9a7a-ecfed7af0a22","order_by":3,"name":"Hongxia Jiang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0klEQVRIiWNgGAWjYBAC9mYgwdjAIMcgAeKyEaGFEarFmAQtDRAisYF4Le3Mzx5+3XE4fcPtHgOGD2WHGfhnNxByGJu5seyZtNwNd84YMM44d5hB4s4BQloYzKQl22xyt93IMWDmbTvMYCCRQEgL+zegFol0M5CWv8RoEWzmMZP82GaTANbCSIwWaWaeMmnGtjTD/TfSCg72nEvnkbhBQAsf//Ftkj/bDstLzkje+OBHmbUc/wwCWkCAmQfKOADEPHgUIgDjD6KUjYJRMApGwYgFAHeGQDXXdBBaAAAAAElFTkSuQmCC","orcid":"","institution":"Zhongnan Hospital of Wuhan University","correspondingAuthor":true,"prefix":"","firstName":"Hongxia","middleName":"","lastName":"Jiang","suffix":""}],"badges":[],"createdAt":"2025-06-08 05:08:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6845563/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6845563/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":85778704,"identity":"4131896a-605f-4bd1-9eef-52d943d431ad","added_by":"auto","created_at":"2025-07-01 14:50:12","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":114420,"visible":true,"origin":"","legend":"\u003cp\u003eData analysis workflow.\u003c/p\u003e","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6845563/v1/4d35632f700eba1f9c9cc7c5.png"},{"id":85779779,"identity":"90569e68-38db-4016-9bf6-8b20d87215a2","added_by":"auto","created_at":"2025-07-01 14:58:12","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":86210,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of important CTA-derived variables and construction of the optimal predictive model for PH. (A) Relationship between the number of variables and model performance: the AUC peaked at 0.89 when 11 features were included; adding more features led to overfitting with a decrease in AUC to 0.87. (B) Importance ranking of the 11 key CTA parameters selected by RFE: right ventricular end-systolic volume (RESV, Gini = 0.32) and main pulmonary artery diameter (PA, Gini = 0.25) were the most critical indicators. (C) The predictive model based on the 11 selected variables was evaluated using 10 machine learning algorithms, including XGBoost, LASSO, and Random Forest. According to AUC, sensitivity, and specificity of the ROC curves, XGBoost was identified as the best-performing model.\u003c/p\u003e","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6845563/v1/cd9ec103b7d3d23d9d4414f6.png"},{"id":85778702,"identity":"9d4ec1e9-8694-4390-adda-d9f5d57b24cf","added_by":"auto","created_at":"2025-07-01 14:50:12","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":81858,"visible":true,"origin":"","legend":"\u003cp\u003eSHAP analysis of the CTA-based predictive model. (A) SHAP beeswarm plot: main pulmonary artery (MPA) diameter made the largest contribution to prediction (SHAP value ±1), while left pulmonary artery (LPA) diameter showed a negative effect. (B) Feature importance ranking: variables were ranked by their impact on PH prediction in descending order, including MPA, LPA, right ventricular stroke volume (RVSV), right ventricular end-systolic volume (RESV), right ventricular output, right ventricular end-diastolic volume (REDV), right pulmonary artery (RPA) diameter, and descending aorta (DAO) diameter at the diaphragmatic level. MPA had the highest mean SHAP value of 0.92. (C) Individualized heatmap: MPA consistently contributed across all samples. (D) Single-sample decision analysis: in high-risk cases, MPA was the primary driving factor.\u003c/p\u003e","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6845563/v1/31f22d1783373f3a7195d5bd.png"},{"id":85779780,"identity":"1f2eaebe-b1ed-4bc6-9f80-91af677adfcb","added_by":"auto","created_at":"2025-07-01 14:58:12","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":137159,"visible":true,"origin":"","legend":"\u003cp\u003eModel performance of the CTA-based predictive model in training and validation cohorts. (A-B) Confusion matrices showed an accuracy of 97.9% in the training set and 90.9% in the validation set. (C-H) In the training cohort, metrics including Accuracy, AUC, F1 score, Precision, Recall, and Specificity demonstrated good model performance. (I-N) In the validation cohort, these same metrics confirmed the model’s stability and reliability.\u003c/p\u003e","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6845563/v1/49a83a9aadb75fc7164807fa.png"},{"id":85778705,"identity":"ebfa8848-2308-45ed-a646-0daa26d18a07","added_by":"auto","created_at":"2025-07-01 14:50:12","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":108034,"visible":true,"origin":"","legend":"\u003cp\u003eIntegration of the multimodal model and decision curve analysis (DCA). (A) Construction workflow of the combined model integrating CTA parameters with laboratory predictors (e.g., sex, age). (B) DCA of laboratory parameter models showed a relatively narrow benefit range. (C) DCA of the integrated model consistently demonstrated higher net benefit than laboratory parameter models. (D-E) Bootstrap validation confirmed the robustness of the integrated model, with AUC values ranging narrowly between 0.88 and 0.92.\u003c/p\u003e","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6845563/v1/b49bb363822d180b9b3932a1.png"},{"id":85778707,"identity":"41b49ec5-45fa-4218-8ed9-b7057bfedf9f","added_by":"auto","created_at":"2025-07-01 14:50:12","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":93096,"visible":true,"origin":"","legend":"\u003cp\u003eDevelopment and validation of the nomogram model. (A) Nomogram composition: sex (13 points for female), age (0–93 points), mean platelet volume (0–70 points), fibrinogen (0–100 points), and CTA index (0–98 points), for a total score of 0–374 points. (B) Calibration curve in the training cohort showed a mean absolute error of 0.048 between predicted and observed risks. (C) Calibration curve in the validation cohort demonstrated slight deviation in the high-score group. (D-E) ROC curves for the training cohort (AUC = 0.998) and validation cohort (AUC = 0.909) demonstrated excellent discriminative ability.\u003c/p\u003e","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-6845563/v1/e2883979be82fdaa2769e038.png"},{"id":98434217,"identity":"d2ae5c64-08bc-49eb-aa7a-ef9c343acc85","added_by":"auto","created_at":"2025-12-17 16:51:43","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1964516,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6845563/v1/42fc290c-52c8-4b00-9041-a684911a5654.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A Novel Clinical Prediction Model for Pulmonary Hypertension Based on Computed Tomography Angiography, Laboratory Data, and basic demographic information","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePulmonary hypertension (PH) is defined as a pathologically elevated mean pulmonary artery pressure (mPAP)\u0026gt;20 mmHg, which is characterized by vascular remodeling, including obstruction, stiffening, and vasoconstriction of the pulmonary vasculature\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]\u003c/sup\u003e. The prevalence of PH varies globally, with approximately 1% of the general population affected, and this rate increases to 10% in individuals over the age of 65 years, leading to a higher risk of mortality\u003csup\u003e[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e. PH is associated with significant morbidity and mortality due to its potential to progress to right heart failure if left untreated\u003csup\u003e[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe gold standard for diagnosing PH is invasive right heart catheterization (RHC), which directly measures hemodynamic parameters such as mPAP and pulmonary vascular resistance (PVR)\u003csup\u003e[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]\u003c/sup\u003e. However, RHC is limited by its invasiveness, high cost, and the risk of complications, making it less accessible for widespread clinical use\u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/sup\u003e. Moreover, RHC is not feasible in patients with contraindications or those residing in remote areas where advanced diagnostic facilities are unavailable\u003csup\u003e[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/sup\u003e. Therefore, there is a pressing need for non-invasive diagnostic tools that can accurately predict PH and guide therapeutic decisions.\u003c/p\u003e \u003cp\u003eIn recent years, advances in medical imaging and machine learning have opened new avenues for the diagnosis and prediction of PH. Computed tomography angiography (CTA) has emerged as a valuable tool for assessing pulmonary vasculature morphology and detecting early signs of vascular remodeling\u003csup\u003e[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/sup\u003e. A lot of information can be gained from CTA, including pulmonary artery diameter, aorta diameter, the artial and ventricular size and function, these infomation reveals the structure and function characteristics of the heart and pulmonary vasculature. Machine learning algorithms, such as gradient boosting decision trees (GBDTs) and logistic regression models, have demonstrated promising results in predicting PH by analyzing CTA-derived parameters and clinical data\u003csup\u003e[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/sup\u003e. These approaches provide a potential alternative to invasive RHC, offering a safer and more accessible method for screening and monitoring PH patients.\u003c/p\u003e \u003cp\u003eThis study aims to develop a novel clinical prediction model for PH combining CTA, laboratory data, as well as basic demographic information such as age and sex/gender, essentially encompassing all the clinically available data with relatively high completeness that we could collect in this study. By leveraging machine learning techniques, we seek to identify key predictors of PH and create a robust model that can accurately diagnose PH. The integration of CTA-derived morphological features with clinical biomarkers represents a significant advancement in the field of PH research, with the potential to improve patient outcomes and reduce the reliance on invasive diagnostic procedures.\u003c/p\u003e"},{"header":"Data Collection and Statistical Analysis","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData Collection\u003c/h2\u003e \u003cp\u003eFrom February 2022 to April 2025, we retrospectively collected clinical and imaging data from patients who underwent pulmonary computed tomography angiography (CTA) at our institution. The study cohort consisted of two groups: Pulmonary Hypertension (PH) Group: Patients with confirmed PH (mean pulmonary arterial pressure\u0026thinsp;\u0026gt;\u0026thinsp;20 mmHg) via right heart catheterization. Control Group: Patients with normal pulmonary CTA findings (no evidence of PH).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eCollected Variables\u003c/h3\u003e\n\u003cp\u003eDemographics (sex, age).\u003c/p\u003e \u003cp\u003eCTA Parameters(Cardiac Morphology: Right atrial enlargement, right ventricular (RV) enlargement, left atrial (LA) enlargement, left ventricular (LV) enlargement, atrial septal defect, ventricular septal defect. Ventricular Dimensions: RV transverse diameter, RV wall thickness, RA superior-inferior diameter, RA left-right diameter, LA anteroposterior diameter, LA superior-inferior diameter, LA left-right diameter, LV transverse diameter.Functional Metrics: LV end-diastolic volume (LEDV), LV end-systolic volume (LESV), LV stroke volume (LSV), LV ejection fraction (LEF), LV mass (Leftmass), LV cardiac output (LeftCO), RV end-diastolic volume (REDV), RV end-systolic volume (RESV), RV stroke volume (RSV), RV ejection fraction (REF), RV cardiac output (Right CO). Vascular Measurements: Pulmonary artery (PA) diameter, left PA diameter (LPA), right PA diameter (RPA), McGoon ratio, Aortic dimensions: Sinotubular junction diameter, ascending aorta diameter, descending aorta diameter, diaphragmatic-level descending aorta diameter ,Total calcium score. )\u003c/p\u003e \u003cp\u003eLaboratory Parameters(Complete Blood Count: White blood cells, red blood cells, hemoglobin, platelets, neutrophil percentage, lymphocyte percentage, monocyte percentage, eosinophil percentage, basophil percentage, absolute neutrophil count, absolute lymphocyte count, absolute monocyte count, absolute eosinophil count, absolute basophil count, hematocrit, mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), red cell distribution width (RDW-CV, RDW-SD), mean platelet volume (MPV). Coagulation Profile: Prothrombin time (PT), international normalized ratio (INR), prothrombin activity, activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen. D-dimer. NT-proBNP. Procalcitonin (PCT). Liver and Renal Function\u0026thinsp;+\u0026thinsp;Electrolytes: Alanine aminotransferase (ALT), aspartate aminotransferase (AST), ALT/AST ratio, direct bilirubin, indirect bilirubin, total bilirubin, total protein, albumin, globulin, albumin-to-globulin ratio (A/G), gamma-glutamyl transferase (GGT), alkaline phosphatase (ALP), total bile acids, creatinine, blood urea nitrogen (BUN), uric acid, CO₂, cystatin C, potassium, sodium, chloride, calcium, magnesium, phosphorus. Lipid Profile: Total cholesterol, triglycerides, high-density lipoprotein (HDL), low-density lipoprotein (LDL), small dense LDL, lipoprotein(a), free fatty acids, phospholipids. Thyroid Function: Free T3, free T4, thyroid-stimulating hormone (TSH). Immunological Markers: Anti-cardiolipin antibodies (IgM, IgG, IgA), complement C3, C4, immunoglobulins (IgG, IgA, IgM, IgE), anti-streptolysin O (ASO), rheumatoid factor (RF), RF antibodies (IgM, IgG, IgA). Arterial Blood Gas Analysis: pH, partial pressure of oxygen (PaO₂), oxygen saturation (SaO₂), partial pressure of CO₂ (PaCO₂), temperature, lactate, standard bicarbonate, actual base excess, standard base excess, anion gap.\u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eData Preprocessing\u003c/p\u003e \u003cp\u003eVariables or samples with \u0026gt;\u0026thinsp;30% missing data were excluded. Remaining missing values were imputed using the k-nearest neighbors (KNN) method. The final dataset included 59 PH patients and 43 controls, with 84 variables retained for analysis.\u003c/p\u003e \u003cp\u003eDescriptive Statistics\u003c/p\u003e \u003cp\u003eContinuous variables were assessed for normality using the Shapiro-Wilk test (p\u0026thinsp;\u0026gt;\u0026thinsp;0.05 considered normally distributed). Normally distributed data: Expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation (SD); compared using independent t-tests (Levene\u0026rsquo;s test for homogeneity of variance, p\u0026thinsp;\u0026gt;\u0026thinsp;0.05) or Welch\u0026rsquo;s t\u0026rsquo;-test (if variances unequal). Non-normally distributed data: Expressed as median (Q1, Q3); compared using the Mann-Whitney U test. Categorical variables: Binary variables: Compared using Pearson\u0026rsquo;s chi-square test. Multiclass variables: Compared using the Kruskal-Wallis test. Statistical significance was set at p\u0026thinsp;\u0026lt;\u0026thinsp;0.05. All analyses were performed using IBM SPSS Statistics 23.\u003c/p\u003e \u003cp\u003eMachine Learning Modeling\u003c/p\u003e \u003cp\u003eData Splitting: Stratified sampling divided the dataset into training (80%) and testing (20%) sets. Feature Selection: Recursive feature elimination (RFE) with 10-fold cross-validation was applied using a random forest-based approach (R caret package). Model Training: Multiple machine learning algorithms were evaluated: LASSO, elastic net, decision tree, random forest, XGBoost, SVM, KNN, naive Bayes, gradient boosting machine (GBM). Model performance was assessed via area under the ROC curve (AUC). Model Interpretability: SHapley Additive exPlanations (SHAP) analysis (Python SHAP package) identified key predictive features. Validation: Bootstrap resampling (1,000 iterations) evaluated model stability. Performance metrics included: Accuracy, AUC, F1-score, precision, recall, specificity.\u003c/p\u003e \u003cp\u003eFinal Predictive Model\u003c/p\u003e \u003cp\u003eUnivariate logistic regression identified significant predictors from clinical/lab variables. These predictors were combined with the optimal CTA-derived features to construct an L1-regularized (LASSO) logistic regression model. A nomogram was developed (R RMS package) to visualize individual risk contributions. Decision curve analysis (DCA) (R RMDA package) assessed clinical utility. Calibration curves and AUC were computed to evaluate model performance in training and test sets.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e1. Baseline Characteristics of the Study Population\u003c/h2\u003e \u003cp\u003eThe overall data analysis workflow of this study has been described in detail (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). A total of 102 patients were ultimately enrolled in this study, including 59 patients (57.8%) with pulmonary hypertension (PH) confirmed by right heart catheterization and 43 controls (42.2%). A comparison of baseline characteristics between the two groups (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) revealed that PH patients had a lower proportion of males (30.5% vs. 60.5%, p\u0026thinsp;=\u0026thinsp;0.003). Regarding imaging parameters, the PH group exhibited typical PH characteristics: the main pulmonary artery diameter was significantly increased (36.64\u0026thinsp;\u0026plusmn;\u0026thinsp;8.7 mm vs. 26.51\u0026thinsp;\u0026plusmn;\u0026thinsp;7 mm, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), right ventricular end-diastolic volume (RVEDV) was markedly enlarged (229.06\u0026thinsp;\u0026plusmn;\u0026thinsp;86.57 mL vs. 173.52\u0026thinsp;\u0026plusmn;\u0026thinsp;63.87 mL, p\u0026thinsp;=\u0026thinsp;0.03), and right ventricular end-systolic volume (RVESV) was significantly higher (123.48 [96.2\u0026ndash;169.88] mL vs. 79.1 [59.38\u0026ndash;88.85] mL, p\u0026thinsp;=\u0026thinsp;0.001). Notably, all cases of right atrial enlargement (43 cases) and right ventricular enlargement (41 cases) were observed exclusively in the PH group (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), consistent with the hemodynamic burden imposed by PH.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline Characteristics of the Variables\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable Name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal (n\u0026thinsp;=\u0026thinsp;102)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eControl Group (n\u0026thinsp;=\u0026thinsp;43)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCase Group (n\u0026thinsp;=\u0026thinsp;59)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e51.46 (33.55-68)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66.74 (51.46\u0026ndash;75.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e42 (30-56.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale gender (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e44(43.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e26(60.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e18(30.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight atrial enlargement, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight ventricular enlargement, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft atrial enlargement, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.012\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft ventricular enlargement, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.082\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAtrial septal defect, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVentricular septal defect, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.019\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft ventricular end-diastolic volume (LVEDV), ml\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e120.07 (97.3-151.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e115.29 (94.92-131.96)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e121.62 (99.94-162.61)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.176\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft ventricular end-systolic volume (LVESV), ml\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e38.25 (27.64\u0026ndash;50.69)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e33.66 (24.27\u0026ndash;44.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41.56 (30.95\u0026ndash;63.72)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.029\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft ventricular stroke volume (LVSV), ml\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e81.36 (66.85\u0026ndash;97.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e81.54 (67.91\u0026ndash;92.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e81.18 (67.1-104.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.595\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft ventricular ejection fraction (LVEF), %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e70 (61\u0026ndash;76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e73 (64.5\u0026ndash;76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e66 (59.5\u0026ndash;75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.053\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft ventricular cardiac output, L/min\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.73 (4.57\u0026ndash;7.58)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.84 (4.56\u0026ndash;7.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.71 (4.62\u0026ndash;8.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.721\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight ventricular end-diastolic volume (RVEDV), ml\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e218.75\u0026thinsp;\u0026plusmn;\u0026thinsp;85.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e173.52\u0026thinsp;\u0026plusmn;\u0026thinsp;63.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e229.06\u0026thinsp;\u0026plusmn;\u0026thinsp;86.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight ventricular end-systolic volume (RVESV), ml\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e112.76 (79.83-159.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e79.1 (59.38\u0026ndash;88.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e123.48 (96.2-169.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight ventricular stroke volume (RVSV), ml\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e89.75 (67.8-124.04)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72.55 (61.85-104.83)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e90.23 (70.71-124.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.282\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight ventricular ejection fraction (RVEF), %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e45.87\u0026thinsp;\u0026plusmn;\u0026thinsp;16.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e47.14\u0026thinsp;\u0026plusmn;\u0026thinsp;17.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e45.57\u0026thinsp;\u0026plusmn;\u0026thinsp;16.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight ventricular cardiac output, L/min\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.28 (4.2\u0026ndash;8.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.49 (3.91\u0026ndash;7.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6.33 (4.64\u0026ndash;8.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.162\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePulmonary artery diameter, mm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e34.45\u0026thinsp;\u0026plusmn;\u0026thinsp;9.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e26.51\u0026thinsp;\u0026plusmn;\u0026thinsp;7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e36.64\u0026thinsp;\u0026plusmn;\u0026thinsp;8.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft pulmonary artery diameter, mm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20.42\u0026thinsp;\u0026plusmn;\u0026thinsp;5.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16.58\u0026thinsp;\u0026plusmn;\u0026thinsp;3.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e21.35\u0026thinsp;\u0026plusmn;\u0026thinsp;5.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight pulmonary artery diameter, mm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22.14\u0026thinsp;\u0026plusmn;\u0026thinsp;7.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16.86\u0026thinsp;\u0026plusmn;\u0026thinsp;5.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e23.42\u0026thinsp;\u0026plusmn;\u0026thinsp;7.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAscending aorta diameter, mm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e30 (24\u0026ndash;33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32 (30\u0026ndash;34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e25 (22.5\u0026ndash;30.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDescending aorta diameter, mm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18 (16\u0026ndash;21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e19.85 (18.25-20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e18 (16\u0026ndash;21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.487\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDiaphragmatic level descending aorta diameter, mm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16.88\u0026thinsp;\u0026plusmn;\u0026thinsp;3.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e17.18\u0026thinsp;\u0026plusmn;\u0026thinsp;4.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e16.81\u0026thinsp;\u0026plusmn;\u0026thinsp;3.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.73\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWhite blood cell count, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.64 (4.86\u0026ndash;7.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.5 (4.93\u0026ndash;7.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.91 (4.8\u0026ndash;6.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.704\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRed blood cell count, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.26 (3.79\u0026ndash;4.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.08 (3.58\u0026ndash;4.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.35 (3.98\u0026ndash;4.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHemoglobin concentration, g/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e127 (110\u0026ndash;144)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e126 (107.05-139.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e133 (113.5-150.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.087\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePlatelet count, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e191.5 (135.5-238.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e213 (173.5\u0026ndash;268)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e175 (117-216.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.006\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNeutrophil percentage, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e62.5\u0026thinsp;\u0026plusmn;\u0026thinsp;11.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e64.4\u0026thinsp;\u0026plusmn;\u0026thinsp;14.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e61.11\u0026thinsp;\u0026plusmn;\u0026thinsp;8.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.15\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLymphocyte percentage, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e27.02\u0026thinsp;\u0026plusmn;\u0026thinsp;10.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25.08\u0026thinsp;\u0026plusmn;\u0026thinsp;12.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e28.43\u0026thinsp;\u0026plusmn;\u0026thinsp;8.38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.11\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMonocyte percentage, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7.8 (6.2\u0026ndash;9.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7.8 (7-9.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7.8 (5.9\u0026ndash;9.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.382\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEosinophil percentage, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.3 (0.7\u0026ndash;2.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.3 (0.7\u0026ndash;2.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.2 (0.7\u0026ndash;2.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.504\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBasophil percentage, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.5 (0.3\u0026ndash;0.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.6 (0.35\u0026ndash;0.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.5 (0.3\u0026ndash;0.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.65\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAbsolute neutrophil count, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.62 (2.79\u0026ndash;4.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.7 (2.81\u0026ndash;5.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.54 (2.72\u0026ndash;4.43)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.173\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAbsolute lymphocyte count, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.5 (1.12\u0026ndash;2.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.37 (1\u0026ndash;2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.6 (1.22\u0026ndash;2.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.099\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAbsolute monocyte count, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.44 (0.37\u0026ndash;0.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.46 (0.4\u0026ndash;0.72)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.43 (0.33\u0026ndash;0.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.147\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAbsolute eosinophil count, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.1 (0.03\u0026ndash;0.14)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.1 (0.04\u0026ndash;0.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.1 (0.02\u0026ndash;0.11)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.186\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAbsolute basophil count, 10⁹/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e47 (46.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23 (53.49)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e24 (40.68)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.288\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHematocrit, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e38.65 (33.73\u0026ndash;43.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e37.3 (32.5\u0026ndash;41.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e40.6 (35.35\u0026ndash;44.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.02\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean corpuscular volume (MCV), fL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e91.3 (87.3\u0026ndash;95.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e91.5 (88.7-95.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e90.65 (85.95\u0026ndash;94.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.266\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean corpuscular hemoglobin (MCH), pg\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e30.3 (28.13\u0026ndash;31.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e30.5 (28.55\u0026ndash;31.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e30 (28.1-31.55)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.272\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean corpuscular hemoglobin concentration (MCHC), g/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e330.5 (321.05\u0026ndash;337)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e332 (323.25\u0026ndash;337)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e330 (317.5-336.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.139\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRed cell distribution width-CV (RDW-CV), %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14 (13.2-15.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13.5 (13.2-15.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14.35 (13.38\u0026ndash;16.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.087\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean platelet volume (MPV), fL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9 (8.1\u0026ndash;10.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.5 (7.95\u0026ndash;9.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9.45 (8.38\u0026ndash;11.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProthrombin time (PT), seconds\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12.1 (11.4\u0026ndash;13.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11.9 (11.15\u0026ndash;13.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e12.2 (11.6\u0026ndash;13)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.255\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInternational Normalized Ratio (INR)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.11 (1.05\u0026ndash;1.21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.09 (1.02\u0026ndash;1.23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.12 (1.06\u0026ndash;1.19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.314\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProthrombin activity, %\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e84 (75\u0026ndash;94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e87 (73-96.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e84 (77-91.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.496\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eActivated partial thromboplastin time (APTT), seconds\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e32.2 (29.9\u0026ndash;34.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32.1 (29.65\u0026ndash;34.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e32.3 (30.18\u0026ndash;34.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.789\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThrombin time (TT), seconds\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14.77\u0026thinsp;\u0026plusmn;\u0026thinsp;1.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14.62\u0026thinsp;\u0026plusmn;\u0026thinsp;1.34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14.88\u0026thinsp;\u0026plusmn;\u0026thinsp;1.38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.34\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFibrinogen concentration, mg/dL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e284 (252.5\u0026ndash;330)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e300 (267\u0026ndash;350)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e275.5 (242-310.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.015\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eD-dimer, ng/mL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e214.5 (101\u0026ndash;760)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e504 (160.5\u0026ndash;1333)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e153 (85\u0026ndash;260)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN-terminal pro-brain natriuretic peptide (NT-proBNP), U/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e279 (88.45\u0026ndash;872.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e195 (72.3\u0026ndash;569)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e375 (103.25\u0026ndash;994)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.136\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlanine aminotransferase (ALT), U/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20 (13\u0026ndash;28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e20 (12\u0026ndash;28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20 (14\u0026ndash;28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.69\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAspartate aminotransferase (AST), U/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22 (18\u0026ndash;29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22 (17\u0026ndash;25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e23 (19\u0026ndash;31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.215\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAST/ALT ratio\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.29 (0.95\u0026ndash;1.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.27 (0.86\u0026ndash;1.77)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.33 (1-1.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.844\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal bilirubin, \u0026micro;mol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e13.8 (9.4-18.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12.1 (8.65\u0026ndash;15.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15.2 (10.3\u0026ndash;20.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.022\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDirect bilirubin, \u0026micro;mol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.7 (2.7\u0026ndash;5.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.4 (2.25\u0026ndash;4.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.9 (2.95\u0026ndash;6.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.138\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIndirect bilirubin, \u0026micro;mol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9.8 (6.2-13.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.5 (6.05\u0026ndash;10.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11.1 (7.1\u0026ndash;14.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.013\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal protein, g/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e67.6 (63.7\u0026ndash;72.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66.8 (61.87\u0026ndash;70.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e68.8 (64.45\u0026ndash;73.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.055\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlbumin, g/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e39.2 (36.53\u0026ndash;41.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e37.65 (34.6-40.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e39.95 (37.42\u0026ndash;42.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.012\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGlobulin, g/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28.3 (25.3\u0026ndash;32.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e28 (25.25\u0026ndash;32.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e28.65 (25.38\u0026ndash;31.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.997\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlbumin/Globulin ratio (A/G ratio)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.33 (1.18\u0026ndash;1.58)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.29 (1.13\u0026ndash;1.52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.34 (1.23\u0026ndash;1.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.194\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGamma-glutamyl transferase (GGT), U/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e27 (16\u0026ndash;58)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22 (13.5\u0026ndash;40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e32 (20\u0026ndash;60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.083\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlkaline phosphatase (ALP), U/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e74 (60.25-95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e80 (60.5\u0026ndash;104)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e72 (60.5\u0026ndash;88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.222\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal bile acid (TBA), \u0026micro;mol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.75 (2.8\u0026ndash;9.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.7 (2-7.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5 (2.95\u0026ndash;12.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.051\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBlood urea nitrogen (BUN), mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.52 (4.4\u0026ndash;6.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.7 (4.28\u0026ndash;8.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.3 (4.43\u0026ndash;6.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.492\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSerum creatinine, \u0026micro;mol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e67.65 (57.05\u0026ndash;82.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e74 (60.65\u0026ndash;89.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e64.7 (53.65\u0026ndash;80.55)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.046\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUric acid, \u0026micro;mol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e343.5 (257-413.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e309 (243.55\u0026ndash;384.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e362.35 (276.62\u0026ndash;483)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.071\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCarbon dioxide (CO₂), mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22.9 (20.7\u0026ndash;25.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23.6 (21.25\u0026ndash;26.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22.8 (20.4-25.05)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.29\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSerum cystatin C, mg/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.99 (0.78\u0026ndash;1.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.01 (0.83\u0026ndash;1.23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.91 (0.77\u0026ndash;1.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.117\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePotassium, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.98\u0026thinsp;\u0026plusmn;\u0026thinsp;0.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.03\u0026thinsp;\u0026plusmn;\u0026thinsp;0.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.94\u0026thinsp;\u0026plusmn;\u0026thinsp;0.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.33\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSodium, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e139.7 (137.4-141.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e139.4 (137.4-141.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e139.85 (137.55-140.78)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.678\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eChloride, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e105.2 (102.75-107.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e104.6 (101.55-107.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e105.3 (103-107.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.245\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalcium, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.27 (2.19\u0026ndash;2.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.23 (2.16\u0026ndash;2.31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.3 (2.21\u0026ndash;2.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMagnesium, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.84 (0.78\u0026ndash;0.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.86 (0.8\u0026ndash;0.92)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.82 (0.78\u0026ndash;0.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.218\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhosphate, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.19 (1.05\u0026ndash;1.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.13 (1-1.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.22 (1.1\u0026ndash;1.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal cholesterol, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.15\u0026thinsp;\u0026plusmn;\u0026thinsp;0.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.9\u0026thinsp;\u0026plusmn;\u0026thinsp;0.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.31\u0026thinsp;\u0026plusmn;\u0026thinsp;0.96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.05\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTriglycerides, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.35 (0.95\u0026ndash;2.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.07 (0.92\u0026ndash;1.77)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.44 (1-2.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.079\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh-density lipoprotein cholesterol (HDL-C), mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.04 (0.88\u0026ndash;1.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.99 (0.84\u0026ndash;1.19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.06 (0.9\u0026ndash;1.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.296\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLow-density lipoprotein cholesterol (LDL-C), mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.47\u0026thinsp;\u0026plusmn;\u0026thinsp;0.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.4\u0026thinsp;\u0026plusmn;\u0026thinsp;0.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.51\u0026thinsp;\u0026plusmn;\u0026thinsp;0.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.55\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSmall dense LDL cholesterol (sdLDL-C), mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.68 (0.48\u0026ndash;1.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.58 (0.46\u0026ndash;0.81)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.88 (0.5\u0026ndash;1.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.068\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLipoprotein(a), mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e89.2 (27-180.05)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e70.45 (16.48-142.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e113.5 (39.75-208.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.137\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFree fatty acids (FFA), \u0026micro;mol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e393.4 (270.76\u0026ndash;605.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e437.5 (360.3-562.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e334.08 (181.2-611.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.153\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePhospholipids, mmol/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.16 (1.89\u0026ndash;2.53)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.02 (1.85\u0026ndash;2.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.36 (1.99\u0026ndash;2.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.002\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\u003eLaboratory test results revealed distinct biochemical characteristics in PH patients: mean platelet volume (MPV) was significantly increased (9.45 [8.38\u0026ndash;11.1] fL vs. 8.5 [7.95\u0026ndash;9.3] fL, p\u0026thinsp;=\u0026thinsp;0.001), fibrinogen levels were decreased (275.5 [242\u0026ndash;310.5] mg/dL vs. 300 [267\u0026ndash;350] mg/dL, p\u0026thinsp;=\u0026thinsp;0.015), and D-dimer levels were markedly elevated (504 [160.5\u0026ndash;1333] ng/mL vs. 153 [85\u0026ndash;260] ng/mL, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), suggesting a significant coagulation dysfunction and prothrombotic state in PH patients. Additionally, albumin levels (39.95 [37.42\u0026ndash;42.75] g/L vs. 37.65 [34.6\u0026ndash;40.98] g/L, p\u0026thinsp;=\u0026thinsp;0.012) and total bilirubin levels (15.2 [10.3\u0026ndash;20.9] \u0026micro;mol/L vs. 12.1 [8.65\u0026ndash;15.95] \u0026micro;mol/L, p\u0026thinsp;=\u0026thinsp;0.022) demonstrated abnormal changes in the PH group, reflecting the pathophysiological alterations of hepatic congestion and systemic inflammation. These findings provided a solid biological basis for the subsequent development of the prediction model.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2. CTA Feature Selection and Model Optimization\u003c/h2\u003e \u003cp\u003eUsing recursive feature elimination (RFE), 11 optimal predictive features were selected from an initial set of 58 CTA parameters. Ten-fold cross-validation confirmed that model performance peaked when using 11 features, achieving an AUC of 0.89 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). As the number of features increased from 5 to 11, the model\u0026rsquo;s performance improved rapidly (AUC from 0.72 to 0.89). However, including more than 11 features led to a decline in AUC to 0.87, indicating a risk of overfitting. This result suggests that these 11 CTA parameters sufficiently captured the imaging characteristics of PH.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe importance ranking of these 11 key CTA parameters was as follows (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB): right ventricular end-systolic volume (RVESV, Gini importance\u0026thinsp;=\u0026thinsp;0.32), left pulmonary artery diameter (LPA, 0.28), main pulmonary artery diameter (PA, 0.25), diaphragmatic-level descending aorta diameter (DHDAD, 0.22), right ventricular end-diastolic volume (RVEDV, 0.19), right ventricular stroke volume (RVSV, 0.18), right pulmonary artery diameter (RPA, 0.17), right cardiac output (RCO, 0.15), ascending aorta diameter (AAD, 0.13), right ventricular ejection fraction (RVEF, 0.11), and descending aorta diameter (DAD, 0.09).\u003c/p\u003e \u003cp\u003eAmong ten machine learning algorithms compared (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC), XGBoost demonstrated superior predictive performance (training set AUC\u0026thinsp;=\u0026thinsp;0.92, sensitivity\u0026thinsp;=\u0026thinsp;86%, specificity\u0026thinsp;=\u0026thinsp;89%), significantly outperforming traditional statistical methods such as LASSO regression (AUC\u0026thinsp;=\u0026thinsp;0.85) and other tree-based models like random forests (AUC\u0026thinsp;=\u0026thinsp;0.88). The advantage of XGBoost likely stems from its gradient boosting mechanism and regularization strategies, which effectively capture complex nonlinear relationships between CTA parameters and PH.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003e3. Interpretability Analysis of the CTA Model\u003c/h3\u003e\n\u003cp\u003eA beeswarm plot visualized the impact of each feature on model output through SHAP values (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). The main pulmonary artery (MPA) demonstrated the widest SHAP value distribution with the largest absolute values (approaching\u0026thinsp;\u0026plusmn;\u0026thinsp;1), confirming its role as the strongest predictive factor. The left pulmonary artery (LPA) showed a distinctive negative effect (primarily distributed in the negative SHAP range). Right heart function parameters (RCO, RVESV, RVEDV) were concentrated in the positive range, consistent with clinical pathophysiology. The density of data points revealed a pronounced rightward shift when MPA exceeded 35 mm, supporting this threshold\u0026rsquo;s clinical relevance.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eA feature importance bar chart based on mean SHAP values (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB) further confirmed this hierarchy: MPA ranked first with a value of 0.92, followed by LPA (0.91, but negative impact), right ventricular systolic parameters (RVSV 0.72, RVESV 0.65) significantly exceeded diastolic parameters (RVEDV 0.51), while RPA (0.28) and DHDAD (0.22) were of secondary importance. This ranking was corroborated by the beeswarm plot, together establishing a robust hierarchy of feature importance. A heatmap of individual feature contributions across 80 cases (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC) demonstrated inter-individual heterogeneity. The MPA (top row) consistently contributed substantially (deep color blocks), LPA (second row) displayed significant negative contributions in approximately one-third of patients (light color blocks), and right heart parameters (RVSV, RVESV, RCO) exhibited diverse contribution patterns, reflecting phenotypic heterogeneity.\u003c/p\u003e \u003cp\u003eA single-sample decision plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD) illustrated the prediction pathway for a specific patient. With a baseline E[f(x)]\u0026thinsp;=\u0026thinsp;1.077 and a predicted f(x)\u0026thinsp;=\u0026thinsp;3.12 (high risk), MPA (+\u0026thinsp;1.149) and RCO (+\u0026thinsp;3.498) were the major positive contributors, while LPA (\u0026ndash;0.842) exerted a significant negative effect. Other parameters contributed in accordance with overall patterns (RVESV\u0026thinsp;+\u0026thinsp;0.716, RPA\u0026thinsp;+\u0026thinsp;1.8), intuitively illustrating the multi-parameter decision logic of the clinical prediction model.\u003c/p\u003e\n\u003ch3\u003e4. CTA Model Validation and Performance Evaluation\u003c/h3\u003e\n\u003cp\u003eThe CTA model exhibited excellent predictive performance in both the training and validation cohorts. In the training cohort, the accuracy for predicting PH was 97.9%, and 97.1% for controls (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). In the validation cohort, the accuracy was 90.9% for PH prediction and 87.5% for controls (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). In the training set (n\u0026thinsp;=\u0026thinsp;82), all evaluation metrics exceeded 0.97: accuracy 97.9%, AUC 0.998 (95% CI: 0.996\u0026ndash;1.000), F1 score 0.978, sensitivity 0.971, and specificity 0.987 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). In the independent validation set (n\u0026thinsp;=\u0026thinsp;20), the model maintained good generalizability: accuracy 90.9%, AUC 0.909 (95% CI: 0.832\u0026ndash;0.986), F1 score 0.875, sensitivity 0.875, and specificity 0.917 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD). ROC curves, precision-recall curves, specificity-sensitivity plots, and accuracy-threshold curves further demonstrated the model\u0026rsquo;s robust performance in both cohorts (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE\u0026ndash;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eL).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e5. Multimodal Model Integration and Optimization\u003c/h2\u003e \u003cp\u003eUnivariate logistic regression identified four independent laboratory predictors: sex, age, MPV, and fibrinogen level. Integrating these factors with the CTA model produced a multimodal prediction model with superior clinical utility (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). Decision curve analysis (DCA) indicated that laboratory model (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB) yielded positive net clinical benefits across a 10\u0026ndash;90% threshold probability range, the integrated model consistently outperformed laboratory model (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC). Notably, within the clinically ambiguous 30\u0026ndash;70% decision threshold, the integrated model identified an additional 15\u0026ndash;18 true PH patients per 100 cases while reducing 10\u0026ndash;12 unnecessary right heart catheterizations, demonstrating substantial clinical value. Bootstrap resampling confirmed the excellent performance of the integrated model (Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD\u0026ndash;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eE).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e6. Nomogram Development and Clinical Application\u003c/h2\u003e \u003cp\u003eA nomogram was developed based on L1-regularized logistic regression incorporating five core predictors (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA). Point allocations and total score ranges were as follows: sex: female, 13 points; male, 0 points. Age: total 0\u0026ndash;93 points (0\u0026ndash;100 years old). MPV: total 0\u0026ndash;70 points (6.5\u0026ndash;13 fL). Fibrinogen: total 0\u0026ndash;100 points (150\u0026ndash;650 mg/dL). CTA index score: total 0\u0026ndash;98 points (range 0\u0026ndash;1). The overall score range was 0 (all optimal values) to 374 (all worst values). Risk stratification was defined as low risk (0\u0026ndash;125, PH probability\u0026thinsp;\u0026lt;\u0026thinsp;30%), intermediate risk (126\u0026ndash;250, PH probability 30\u0026ndash;70%), and high risk (251\u0026ndash;374, PH probability\u0026thinsp;\u0026gt;\u0026thinsp;70%). Calibration curves showed good agreement in both training (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB) and validation cohorts (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC), with an overall mean absolute error of only 0.048, indicating excellent calibration. The model exhibited outstanding discriminatory ability in the training set (AUC\u0026thinsp;=\u0026thinsp;0.998) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eD) and remained stable in the validation set (AUC\u0026thinsp;=\u0026thinsp;0.909) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eE).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eA clinical application example: a 72-year-old female with an MPV of 12.8 fL, fibrinogen of 180 mg/dL, and a CTA index score of 0.82, achieved a total score of 309 (13 [sex]\u0026thinsp;+\u0026thinsp;67 [age]\u0026thinsp;+\u0026thinsp;63 [MPV]\u0026thinsp;+\u0026thinsp;86 [fibrinogen]\u0026thinsp;+\u0026thinsp;80 [CTA index]\u0026thinsp;=\u0026thinsp;309). This score (\u0026gt;\u0026thinsp;250) classified her as high risk with a PH probability exceeding 90%. Clinical recommendations included immediate right heart catheterization and initiation of targeted therapy. This tool is especially valuable for PH screening and referral decision-making in primary healthcare settings.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eConstructing of a novel clinical prediction model for the diagnosis of PH based on CTA, laboratory data, and basic demographic information represents a significant breakthrough in the diagnosis and management of this complex condition. Our study leverages advanced machine learning techniques to identify optimal predictors of PH from a comprehensive dataset encompassing both imaging and clinical variables, which nearly encompasses all clinically collectible valuable information.The integration of these diverse data sources enables a more holistic understanding of PH pathophysiology and enhances the accuracy of predictive modeling.\u003c/p\u003e \u003cp\u003eOur findings indicate that CTA-derived parameters such as RESV, LPA, PA, DHDAD, REDV, RSV, RPA, RCO, AAD, REF, and DAD are critical predictors of PH. These morphological features reflect the extent of pulmonary vascular remodeling, which is a hallmark of PH pathophysiology\u003csup\u003e[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]\u003c/sup\u003e. Selecting XGBOOST as the optimal machine learning model underscores the importance of ensemble methods in capturing complex interactions between variables\u003csup\u003e[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/sup\u003e. Furthermore, the inclusion of clinical variables such as sex, age, mean platelet volume, and fibrogen content in the final model highlights the multifactorial nature of PH and the need for a comprehensive approach for risk assessment\u003csup\u003e[\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe performance metrics of our model, including accuracy, AUC, F1 score, precision, recall, and specificity, demonstrate exceptional predictive capabilities. Specifically, the model achieved an AUC of 0.998 in the training set and 0.909 in the validation set, indicating high discriminatory power\u003csup\u003e[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]\u003c/sup\u003e. This level of performance is comparable to or even superior to existing models that rely solely on clinical or imaging data\u003csup\u003e[\u003cspan additionalcitationids=\"CR16\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]\u003c/sup\u003e. Moreover, the calibration curve analysis confirms that our model maintains its predictive validity across different patient populations, ensuring reliability in clinical practice .\u003c/p\u003e \u003cp\u003eOne of the key strengths of our study lies in the rigorous methodology employed to select and validate predictors. Recursive feature elimination (RFE) based on random forests was used to identify the most relevant features from the vast CTA dataset, followed by cross-validation and SHAP analysis to ensure robustness and interpretability\u003csup\u003e[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]\u003c/sup\u003e. This systematic approach minimizes overfitting and ensures that the final model generalizes well to new data. Another highlight of our study is that all patients in the PH groups were confirmed to have pulmonary hypertension via RHC (the gold standard for PH diagnosis), while all control subjects showed no significant structural or functional abnormalities in the cardiovascular system based on CTA findings. The zero misdiagnosis rate in both the case and control groups further enhances the credibility of the final model.\u003c/p\u003e \u003cp\u003eThe practical implications of our model are substantial. For patients who are contraindicated for RHC or reside in remote areas without access to advanced diagnostic facilities, our model provides a reliable alternative for PH screening. Additionally, the model can inform therapeutic decisions by identifying high-risk patients who require urgent intervention or those who may benefit from targeted treatments such as endothelin receptor antagonists or phosphodiesterase inhibitors .\u003c/p\u003e \u003cp\u003eHowever, several limitations must be acknowledged. The retrospective nature of our study design introduces potential biases related to data collection and patient selection. Future prospective studies are necessary to validate our findings in diverse populations and settings. Furthermore, while our model demonstrates high accuracy, it may not capture all nuances of PH pathophysiology, particularly in cases where comorbidities or genetic factors play a significant role\u003csup\u003e[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]\u003c/sup\u003e .\u003c/p\u003e \u003cp\u003eIn conclusion, our study presents a novel clinical prediction model for PH that integrates CTA-derived morphological features with laboratory data and basic demographic information. This model represents a significant advancement in non-invasive PH diagnosis and has the potential to improve patient outcomes by facilitating early detection and personalized treatment strategies. Future research should focus on validating this model in larger, more diverse populations and exploring its applicability to other cardiovascular conditions.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAUC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eArea under the receiver operating characteristic curve\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eALT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eAlanine aminotransferase\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAPTT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eActivated partial thromboplastin time\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAST\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eAspartate aminotransferase\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCTA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eComputed tomography angiography\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDCA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDecision curve analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eHDL\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eHigh-density lipoprotein\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eINR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eInternational normalized ratio\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eLDL\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLow-density lipoprotein\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eLPA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLeft pulmonary artery\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eLVEDV\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLeft ventricular end-diastolic volume\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eLVEF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLeft ventricular ejection fraction\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eLVESV\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLeft ventricular end-systolic volume\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMPA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMain pulmonary artery\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMPV\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMean platelet volume\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNT-proBNP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eN-terminal pro-brain natriuretic peptide\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePulmonary artery\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePH\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePulmonary hypertension\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eProthrombin time\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRCO\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRight cardiac output\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eREDV\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRight ventricular end-diastolic volume\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eREF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRight ventricular ejection fraction\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRESV\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRight ventricular end-systolic volume\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRFE\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRecursive feature elimination\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRHC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRight heart catheterization\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRPA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRight pulmonary artery\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRSV\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRight ventricular stroke volume\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRV\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRight ventricle/ventricular\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRVEF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRight ventricular ejection fraction\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSHAP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSHapley Additive exPlanations\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eTT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eThrombin time.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eEthics statement\u003c/h2\u003e\n\u003cp\u003eThis study was conducted in accordance with the ethical principles of the Declaration of Helsinki and was approved by the Medical Ethics Committee of Zhongnan Hospital, Wuhan University. All data were anonymized to protect patient confidentiality, and only de-identified information was used for analysis. Due to the retrospective nature of the study, the need to obtain the informed consent was waived by the Medical Ethics Committee of Zhongnan Hospital, Wuhan University.\u003c/p\u003e\n\u003ch2\u003eCompeting Interests\u003c/h2\u003e\n\u003cp\u003eAll the authors declare that no conflict of interest exists.\u003c/p\u003e\n\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eThis study was funded by the following programs: The National Natural Science Foundation of China (No.82300078), The Fundamental Research Funds for the Central Universities (No.2042023kf0059), Wuhan University Clinical Medicine\u0026thinsp;+\u0026thinsp;Youth Supporting Program (No.413000557).\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eAll authors contributed to this article: Qian Cheng contributed to Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing\u0026mdash;original draft. Xiaojun Hao contributed to Data curation, Formal analysis, Validation, Investigation. Hao Li contributed to Formal analysis, Validation, Investigation, Visualization, Methodology. Hongxia Jiang contributed to Resources, Supervision, Funding acquisition, Project administration, Writing\u0026mdash;review \u0026amp; editing, and verifing the underlying data reported in the manuscript. All authors have read and agreed to the published version of the article.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eAll data are available from the corresponding author upon reasonable request\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSimonneau, G. et al. Haemodynamic definitions and updated clinical classification of pulmonary hypertension. \u003cem\u003eEur. Respir. J.\u003c/em\u003e \u003cb\u003e53\u003c/b\u003e (1), 1801913 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGali\u0026egrave;, N. et al. \u003cem\u003e2015 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension\u003c/em\u003e. \u003cem\u003eEur. Respir. J.\u003c/em\u003e \u003cb\u003e46\u003c/b\u003e(4): pp. 903\u0026ndash;975 .\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKularatne, M. et al. \u003cem\u003eUpdated Clinical Classification and Hemodynamic Definitions of Pulmonary Hypertension and Its Clinical Implications\u003c/em\u003e. \u003cem\u003eJ. Cardiovasc. Dev. Dis.\u003c/em\u003e, \u003cb\u003e11\u003c/b\u003e(3). (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonzalez-Hermosillo, L. M. et al. Right Heart Catheterization (RHC): A Comprehensive Review of Provocation Tests and Hepatic Hemodynamics in Patients With Pulmonary Hypertension (PH). \u003cem\u003eCurr. Probl. Cardiol.\u003c/em\u003e \u003cb\u003e47\u003c/b\u003e (12), 101351 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, M. T. et al. Application of homocysteine as a non-invasive and effort-free measurements for risk assessment of patients with pulmonary hypertension. \u003cem\u003eCardiol. J.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e (2), 285\u0026ndash;299 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingh, V., Dwivedi, S. N. \u0026amp; Deo, S. V. S. Ordinal logistic regression model describing factors associated with extent of nodal involvement in oral cancer patients and its prospective validation. \u003cem\u003eBMC Med. Res. Methodol.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e (1), 95 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFisher, M. R. et al. Accuracy of Doppler echocardiography in the hemodynamic assessment of pulmonary hypertension. \u003cem\u003eAm. J. Respir Crit. Care Med.\u003c/em\u003e \u003cb\u003e179\u003c/b\u003e (7), 615\u0026ndash;621 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan den Eynde, J. et al. Successfully implemented artificial intelligence and machine learning applications in cardiology: State-of-the-art review. \u003cem\u003eTrends Cardiovasc. Med.\u003c/em\u003e \u003cb\u003e33\u003c/b\u003e (5), 265\u0026ndash;271 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSpruijt, O. A. et al. Predicting pulmonary hypertension with standard computed tomography pulmonary angiography. \u003cem\u003eInt. J. Cardiovasc. Imaging\u003c/em\u003e. \u003cb\u003e31\u003c/b\u003e (4), 871\u0026ndash;879 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, Q. Y. et al. Development and Validation of a Machine-Learning Model for Prediction of Extubation Failure in Intensive Care Units. \u003cem\u003eFront. Med. (Lausanne)\u003c/em\u003e. \u003cb\u003e8\u003c/b\u003e, 676343 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoris, P., Melazzini, F. \u0026amp; Balduini, C. L. New roles for mean platelet volume measurement in the clinical practice? \u003cem\u003ePlatelets\u003c/em\u003e \u003cb\u003e27\u003c/b\u003e (7), 607\u0026ndash;612 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSokou, R. et al. \u003cem\u003eDiagnostic and Prognostic Value of Hematological Parameters in Necrotizing Enterocolitis: A Systematic Review\u003c/em\u003e. \u003cem\u003eJ. Clin. Med.\u003c/em\u003e, \u003cb\u003e14\u003c/b\u003e(7). (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUpadhyay, R. K. \u003cem\u003eEmerging risk biomarkers in cardiovascular diseases and disorders.\u003c/em\u003e J Lipids, 2015: p. 971453. (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, N. et al. \u003cem\u003eMachine Learning Based on Computed Tomography Pulmonary Angiography in Evaluating Pulmonary Artery Pressure in Patients with Pulmonary Hypertension\u003c/em\u003e. \u003cem\u003eJ. Clin. Med.\u003c/em\u003e, \u003cb\u003e12\u003c/b\u003e(4). (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeng, Y. et al. A novel clinical prediction scoring system of high-altitude pulmonary hypertension. \u003cem\u003eFront. Cardiovasc. Med.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 1290895 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLui, J. K. et al. A Multimodal Prediction Model for Diagnosing Pulmonary Hypertension in Systemic Sclerosis. \u003cem\u003eArthritis Care Res. (Hoboken)\u003c/em\u003e. \u003cb\u003e75\u003c/b\u003e (7), 1462\u0026ndash;1468 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNathan, S. D. et al. Derivation and validation of a noninvasive prediction tool to identify pulmonary hypertension in patients with IPF: Evolution of the model FORD. \u003cem\u003eJ. Heart Lung Transpl.\u003c/em\u003e \u003cb\u003e43\u003c/b\u003e (4), 547\u0026ndash;553 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEnde-Verhaar, Y. M. et al. Usefulness of standard computed tomography pulmonary angiography performed for acute pulmonary embolism for identification of chronic thromboembolic pulmonary hypertension: results of the InShape III study. \u003cem\u003eJ. Heart Lung Transpl.\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e (7), 731\u0026ndash;738 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShahin, Y. et al. Quantitative CT Evaluation of Small Pulmonary Vessels Has Functional and Prognostic Value in Pulmonary Hypertension. \u003cem\u003eRadiology\u003c/em\u003e \u003cb\u003e305\u003c/b\u003e (2), 431\u0026ndash;440 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArvanitaki, A. et al. \u003cem\u003eNoninvasive diagnostic modalities and prediction models for detecting pulmonary hypertension associated with interstitial lung disease: a narrative review\u003c/em\u003e. \u003cem\u003eEur. Respir Rev.\u003c/em\u003e, \u003cb\u003e33\u003c/b\u003e(174). (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePalazzini, M. et al. Pulmonary hypertension due to left heart disease: analysis of survival according to the haemodynamic classification of the 2015 ESC/ERS guidelines and insights for future changes. \u003cem\u003eEur. J. Heart Fail.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e (2), 248\u0026ndash;255 (2018).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Pulmonary hypertension, machine learning, predictive model","lastPublishedDoi":"10.21203/rs.3.rs-6845563/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6845563/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePulmonary hypertension (PH) is defined as mean pulmonary artery pressure (mPAP)\u0026thinsp;\u0026gt;\u0026thinsp;20 mmHg and diagnosed invasively via right heart catheterization (RHC). In this study, we developed a noninvasive PH prediction model. Recursive feature elimination (RFE) selected 11 CTA indicators (RESV, LPA, PA, DHDAD, REDV, RSV, RPA, RCO, AAD, REF, and DAD) as predictors. Among ten machine learning models, XGBoost performed best, with SHAP analysis highlighting MPA as the most influential variable. The CTA model achieved high accuracy (training: 97.9%, validation: 90.9%) and robust metrics (AUC\u0026thinsp;\u0026gt;\u0026thinsp;0.875). Univariate logistic regression identified additional predictors (sex, age, platelet volume, fibrinogen), which, combined with CTA data, improved performance (AUC: training 0.998, validation 0.909). The final logistic regression model with L1 regularization was visualized as a nomogram. Decision curve analysis confirmed clinical utility. This noninvasive approach, integrating CTA, lab tests, and demographics, aids PH diagnosis in RHC-contraindicated or resource-limited settings.\u003c/p\u003e","manuscriptTitle":"A Novel Clinical Prediction Model for Pulmonary Hypertension Based on Computed Tomography Angiography, Laboratory Data, and basic demographic information","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-01 14:50:07","doi":"10.21203/rs.3.rs-6845563/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":"6363eba1-d995-48bb-89a2-2c86b2f095bf","owner":[],"postedDate":"July 1st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":50713410,"name":"Health sciences/Medical research/Experimental models of disease"},{"id":50713411,"name":"Health sciences/Medical research/Pre clinical studies"}],"tags":[],"updatedAt":"2025-12-16T04:53:56+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-01 14:50:07","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6845563","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6845563","identity":"rs-6845563","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}