CT-Based Radiomics Model for Preoperative Prediction of Histological Grade and Postoperative Survival in Hepatocellular Carcinoma

CT-Based Radiomics Model for Preoperative Prediction of Histological Grade and Postoperative Survival in Hepatocellular Carcinoma | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article CT-Based Radiomics Model for Preoperative Prediction of Histological Grade and Postoperative Survival in Hepatocellular Carcinoma Mingxing Sun, Zhijun Wang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9078107/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 Objective To develop and validate a computed tomography (CT)-based radiomics model for predicting histological grade and postoperative survival in patients with hepatocellular carcinoma (HCC). Methods This retrospective study included 385 patients with pathologically confirmed HCC who underwent preoperative CT examinations between January 2013 and September 2022. Patients were randomly assigned to a training cohort (n = 265) and a testing cohort (n = 120). Radiomics features were extracted from portal venous phase CT images using standardized radiomics pipelines. Feature selection was performed using intraclass correlation coefficient (ICC) analysis and least absolute shrinkage and selection operator (LASSO) regression. Model performance was evaluated using receiver operating characteristic (ROC) analysis. Survival outcomes were assessed using Kaplan–Meier analysis and Cox proportional hazards regression. Results Among 396 extracted radiomics features, 180 features demonstrated good reproducibility (ICC > 0.75). After dimensionality reduction, seven radiomics features were selected to construct the radiomics signature. The model achieved an AUC of 0.889 in the training cohort and 0.941 in the testing cohort. Survival analysis identified the radiomics feature Dif.Scale1.2 as an independent predictor of overall survival (HR = 0.319, 95% CI: 0.143–0.701, P = 0.002). Conclusion CT-based radiomics may serve as a promising noninvasive imaging biomarker for predicting tumor differentiation and postoperative survival in hepatocellular carcinoma. Radiomics Computed tomography Hepatocellular carcinoma Tumor differentiation Prognosis Survival prediction Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1 Introduction Hepatocellular carcinoma (HCC) is one of the most common primary liver malignancies and remains a major cause of cancer-related mortality worldwide. Tumor differentiation grade reflects tumor aggressiveness and has important implications for treatment planning and prognosis. Poorly differentiated HCC is associated with increased tumor invasiveness, early recurrence, and poor survival outcomes. Contrast-enhanced CT is widely used for the diagnosis and staging of HCC because of its accessibility and high spatial resolution. However, conventional CT interpretation relies primarily on qualitative image assessment and may not adequately capture intratumoral heterogeneity. Radiomics refers to the high-throughput extraction of quantitative imaging features from medical images using advanced computational algorithms. These features may reflect tumor heterogeneity and biological behavior. Recent studies have explored the role of radiomics in predicting tumor grade and clinical outcomes in hepatocellular carcinoma. However, the potential value of CT-based radiomics for simultaneously predicting tumor differentiation and postoperative survival remains insufficiently investigated. Therefore, the aim of this study was to develop and validate a CT-based radiomics model for predicting histological grade and postoperative survival in patients with hepatocellular carcinoma. 2 Materials and Methods 2.1 Patient Population This retrospective study was approved by the Institutional Review Board of Ningxia Medical University General Hospital. The requirement for informed consent was waived due to the retrospective nature of the study. A total of 385 patients with pathologically confirmed hepatocellular carcinoma who underwent preoperative CT examinations between January 2013 and September 2022 were included. Tumor differentiation was classified according to the Edmondson–Steiner grading system. Grade I–II tumors were considered low-grade HCC, whereas grade III–IV tumors were classified as high-grade HCC. Patients were randomly assigned to a training cohort (n = 265) and a testing cohort (n = 120) (Table 1 ). A subgroup of 60 patients who underwent hepatectomy with complete follow-up data was included for survival analysis. Table 1 Clinical characteristics of patients in the training and testing cohorts. Feature Training Cohort (n = 265) Testing Cohort (n = 120) Total (n = 385) Age (years) Mean ± SD Mean ± SD Mean ± SD Gender Male: 103, Female: 32 Male: 46, Female: 16 Male: 149, Female: 48 AFP (ng/mL) Median (Q1, Q3) Median (Q1, Q3) Median (Q1, Q3) Tumor Grade Low-grade: 135, High-grade: 130 Low-grade: 55, High-grade: 65 Low-grade: 190, High-grade: 195 Rad-score Mean ± SD Mean ± SD Mean ± SD 2.2 CT Image Acquisition All CT examinations were performed using a 256-slice spiral CT scanner (Brilliance iCT, Philips Healthcare). Portal venous phase images were used for radiomics analysis because they provide optimal tumor-to-liver contrast for lesion delineation. Images were reconstructed with a slice thickness of ≤ 2 mm. 2.3 Tumor Segmentation Tumor segmentation was performed using ITK-SNAP software. Regions of interest were manually delineated along the tumor boundary on each CT slice by two experienced radiologists. Three-dimensional volumes of interest were subsequently generated. 2.4 Radiomics Feature Extraction Radiomics features were extracted according to the Image Biomarker Standardisation Initiative (IBSI) guidelines. A total of 396 radiomics features were extracted, including: • shape features • first-order histogram features • texture features derived from gray-level matrices 2.5 Feature Selection and Model Construction Feature reproducibility was assessed using ICC analysis. Features with ICC > 0.75 were retained. LASSO regression with ten-fold cross-validation was applied to reduce feature dimensionality and prevent overfitting. Seven optimal features were selected to construct the radiomics signature (Table 2 ). Table 2 Radiomics feature selection process. Feature Type Feature Name Description Shape Features Volume Volume of the tumor (cm³) Surface Area Surface area of the tumor (cm²) Sphericity Sphericity of the tumor First-Order Features Mean Intensity Average pixel intensity in ROI Standard Deviation Standard deviation of intensity Skewness Skewness of pixel intensity Texture Features Homogeneity Homogeneity of the tumor texture Entropy Entropy of the tumor texture Correlation Correlation of pixel intensities 2.6 Model Evaluation The predictive performance of the radiomics model was evaluated using ROC analysis. 2.7 Survival Analysis Overall survival was defined as the interval from surgery to death or last follow-up. Disease-free survival was defined as the interval from surgery to tumor recurrence or death. Kaplan–Meier survival curves were generated and Cox regression analysis was used to identify independent prognostic factors (Table 3 ). Table 3 Multivariable Cox regression analysis for overall survival. Feature Hazard Ratio (HR) 95% Confidence Interval (CI) P-value Dif.Scale1.2 0.319 0.143–0.701 0.002 Dif.Scale1.3 0.432 0.227–0.832 0.017 Dif.Scale1.4 0.526 0.238–0.818 0.045 Text.Homogeneity 1.202 0.983–1.459 0.073 Text.Entropy 1.087 0.929–1.264 0.295 Shape.Volume 1.001 0.997–1.005 0.432 3 Results The radiomics score was significantly higher in patients with high-grade HCC than in those with low-grade tumors (P < 0.001). The radiomics workflow is illustrated in Fig. 1. The categories of extracted radiomics features are shown in Fig. 2. The evaluation of the radiomics signature is presented in Fig. 3. The model achieved an AUC of 0.889 in the training cohort and 0.941 in the testing cohort (Fig. 4). Prognostic Analysis Among the 60 patients included in survival analysis, 40 experienced recurrence and 22 died during follow-up. Cox regression analysis identified Dif.Scale1.2 as an independent predictor of overall survival (Table 4 ). Kaplan–Meier curves for overall survival are shown in Fig. 5. Table 4 Prognostic performance of the radiomics model. Variable Hazard Ratio (HR) 95% Confidence Interval P value Age 1.12 0.94–1.32 0.215 Tumor size 1.48 1.05–2.08 0.024 AFP level 1.31 0.97–1.76 0.078 Radiomics feature (Dif.Scale1.2) 0.319 0.143–0.701 0.002 Kaplan–Meier curves for disease-free survival are shown in Fig. 6. 4 Discussion This study developed and validated a CT-based radiomics model for predicting histological grade and postoperative survival in hepatocellular carcinoma. The results demonstrate that radiomics features derived from CT images can effectively differentiate tumor grade and provide prognostic information. Tumor heterogeneity plays an important role in tumor progression and treatment response. Radiomics enables quantitative characterization of tumor heterogeneity from routine medical images. Our findings are consistent with previous studies demonstrating the predictive value of radiomics in hepatocellular carcinoma. Several limitations should be acknowledged. First, this was a retrospective single-center study. Second, the survival cohort was relatively small. Third, external validation was not performed. Future multicenter studies with larger sample sizes are warranted to further validate these findings. Therefore, CT-based radiomics may provide a useful imaging biomarker for individualized risk stratification in patients with hepatocellular carcinoma. 5 Conclusion CT-based radiomics may serve as a promising noninvasive imaging biomarker for predicting tumor differentiation and postoperative survival in hepatocellular carcinoma. Declarations Ethics approval and consent to participate This retrospective study was approved by the Institutional Review Board of Ningxia Medical University General Hospital (Ethics approval number: KYLL-2024-0665). The requirement for informed consent was waived due to the retrospective nature of the study. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests. Author Contribution Sun Mingxing performed the data collection, radiomics analysis, statistical analysis, and drafted the manuscript. Wang Zhijun conceived and supervised the study and critically revised the manuscript. Both authors read and approved the final manuscript. Data Availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. References Lambin P, Rios-Velazquez E, Leijenaar R, Carvalho S, van Stiphout RGPM, Granton P, et al. Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer. 2012;48:441–6. Gillies RJ, Kinahan PE, Hricak H. Radiomics: images are more than pictures, they are data. Radiology. 2016;278:563–77. Aerts HJWL, Velazquez ER, Leijenaar RTH, Parmar C, Grossmann P, Carvalho S, et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat Commun. 2014;5:4006. Tibshirani R. Regression shrinkage and selection via the LASSO. J R Stat Soc B. 1996;58:267–88. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology. 1982;143:29–36. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457–81. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide. CA Cancer J Clin. 2018;68:394–424. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide. CA Cancer J Clin. 2021;71:209–49. Llovet JM, Kelley RK, Villanueva A, Singal AG, Pikarsky E, Roayaie S, et al. Hepatocellular carcinoma. Nat Rev Dis Primers. 2021;7:6. El-Serag HB. Epidemiology of hepatocellular carcinoma. Gastroenterology. 2012;142:1264–73. Ronot M, Vilgrain V. Hepatocellular carcinoma: diagnostic criteria and imaging techniques. Radiology. 2016;279:324–40. European Association for the Study of the Liver. EASL Clinical Practice Guidelines: management of hepatocellular carcinoma. J Hepatol. 2018;69:182–236. Kudo M. Imaging diagnosis of hepatocellular carcinoma and premalignant/borderline lesions. Liver Cancer. 2013;2:365–77. Limkin EJ, Sun R, Dercle L, Zacharaki EI, Robert C, Reuzé S, et al. Promises and challenges for the implementation of computational medical imaging (radiomics) in oncology. Nat Rev Clin Oncol. 2017;14:749–62. Yip SSF, Aerts HJWL. Applications and limitations of radiomics. Phys Med Biol. 2016;61:R150–66. Mao B, Zhang L, Ning P, Ding F, Wu F, Lu G, et al. Preoperative prediction for pathological grade of hepatocellular carcinoma via machine learning-based radiomics. Eur Radiol. 2020;30:6924–32. Huang YQ, Liang CH, He L, Tian J, Liang CS, Chen X, et al. Development and validation of a radiomics nomogram for preoperative prediction of microvascular invasion in hepatocellular carcinoma. Radiology. 2019;292:573–80. Xu X, Zhang HL, Liu QP, Sun SW, Zhang J, Zhu FP, et al. Radiomic analysis of contrast-enhanced CT predicts microvascular invasion in hepatocellular carcinoma. J Hepatol. 2019;70:1133–44. Edmondson HA, Steiner PE. Primary carcinoma of the liver: a study of 100 cases among 48,900 necropsies. Cancer. 1954;7:462–503. Chernyak V, Fowler KJ, Kamaya A, Kielar AZ, Elsayes KM, Bashir MR, et al. LI-RADS version 2018 imaging algorithm. Radiology. 2018;289:816–30. Yushkevich PA, Piven J, Hazlett HC, Smith RG, Ho S, Gee JC, et al. User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. NeuroImage. 2006;31:1116–28. Zwanenburg A, Vallières M, Abdalah MA, Aerts HJWL, Andrearczyk V, Apte A, et al. The image biomarker standardisation initiative. Radiology. 2020;295:328–38. van Griethuysen JJM, Fedorov A, Parmar C, Hosny A, Aucoin N, Narayan V, et al. Computational radiomics system to decode the radiographic phenotype. Cancer Res. 2017;77:e104–7. Kumar V, Gu Y, Basu S, Berglund A, Eschrich SA, Schabath MB, et al. Radiomics: the process and the challenges. Magn Reson Imaging. 2012;30:1234–48. Wu M, Tan H, Gao F, Hai J, Ning P, Chen J, et al. Predicting the grade of hepatocellular carcinoma using radiomics. Front Oncol. 2020;10:177. Lubner MG, Smith AD, Sandrasegaran K, Sahani DV, Pickhardt PJ. CT texture analysis: definitions, applications, biologic correlates, and challenges. Radiographics. 2017;37:1483–503. Zwanenburg A. Radiomics in nuclear medicine imaging. Eur J Nucl Med Mol Imaging. 2019;46:2638–46. Parekh VS, Jacobs MA. Radiomics: a new application from established techniques. Expert Rev Precis Med Drug Dev. 2016;1:207–26. Sun R, Limkin EJ, Vakalopoulou M, Dercle L, Champiat S, Han SR, et al. A radiomics approach to assess tumour heterogeneity. Nat Rev Clin Oncol. 2020;17:757–72. Parmar C, Grossmann P, Bussink J, Lambin P, Aerts HJWL. Machine learning methods for quantitative radiomic biomarkers. Nat Rev Clin Oncol. 2015;12:104–17. 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-9078107","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":609923822,"identity":"9ac89bcb-bc5c-48ce-a4bc-8b441acd88ca","order_by":0,"name":"Mingxing Sun","email":"","orcid":"","institution":"Ningxia Medical University General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Mingxing","middleName":"","lastName":"Sun","suffix":""},{"id":609923823,"identity":"fa86dd17-8c08-45c5-a2b8-d0beee9283d8","order_by":1,"name":"Zhijun Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/0lEQVRIie3RsUoDMQDG8YTAueS89UJ9iEggtlDaV7lycNOhuHWSSOVW1+uLlI45DuwS7Ro4B8VJ6JDb7NbYLl2a6yiY//ANIb8hBACf768GRYwAQLL+mQ5xFImzSTD5Aiq7IqU8i/wuZgw+1UMqEvdtunqtv7fL/gVtBO/dizWmQELT5g6ibrM+UTGi7zLrzZcNvkECkfniNOEy5/S6sEQnL3GoGjwQMkChi6w3nE72xG5YvGEqkw6ic/ZR7UmKWFjIbjLWGw4fLSE6g5+lSjEpq5nzLeQ5Z+22eEgv9Z2RZjoaR9GsMq2D2ILYTnp0cPgmV8jYGXXd8vl8vn/cDtaIV+nKfdIOAAAAAElFTkSuQmCC","orcid":"","institution":"Ningxia Medical University General Hospital","correspondingAuthor":true,"prefix":"","firstName":"Zhijun","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2026-03-10 02:24:03","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9078107/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9078107/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105565609,"identity":"99956408-67d2-4e40-98f2-8a5336f6f3e3","added_by":"auto","created_at":"2026-03-27 12:53:46","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":590970,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend\u003c/p\u003e","description":"","filename":"figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-9078107/v1/e0434f98746da5f0c650c213.png"},{"id":106092949,"identity":"bee973ba-a5ac-482a-af77-8dc12d491be5","added_by":"auto","created_at":"2026-04-03 11:31:21","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1762151,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend\u003c/p\u003e","description":"","filename":"figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-9078107/v1/997c66a9a790fbbf51e121d8.png"},{"id":105564708,"identity":"5fdd64fc-09bf-498c-8fe3-a2e4fdcb1b45","added_by":"auto","created_at":"2026-03-27 12:50:34","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2246793,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend\u003c/p\u003e","description":"","filename":"figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-9078107/v1/1bde8ae2fc318bb6a551ad0c.png"},{"id":105363582,"identity":"ca8ae8de-f6b5-4964-8a7a-88aee5592d3a","added_by":"auto","created_at":"2026-03-25 08:20:50","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":365863,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-9078107/v1/56cae31f168c27b62cad1406.png"},{"id":105565342,"identity":"f303da9b-8937-4b4a-b585-86801831eab1","added_by":"auto","created_at":"2026-03-27 12:52:58","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":180669,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-9078107/v1/18b1a00a3c51ea9f8d98507d.png"},{"id":105565389,"identity":"f76a7c88-5e00-4dde-8749-d3e4a0af4878","added_by":"auto","created_at":"2026-03-27 12:53:06","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":194262,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-9078107/v1/d179270bd4632cf826baa15c.png"},{"id":108760032,"identity":"37f3e484-b7ba-4a24-b076-a79cfec792dd","added_by":"auto","created_at":"2026-05-08 06:26:15","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4586579,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9078107/v1/6160a72c-3bf2-4652-86de-44f6275715d7.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"CT-Based Radiomics Model for Preoperative Prediction of Histological Grade and Postoperative Survival in Hepatocellular Carcinoma","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eHepatocellular carcinoma (HCC) is one of the most common primary liver malignancies and remains a major cause of cancer-related mortality worldwide. Tumor differentiation grade reflects tumor aggressiveness and has important implications for treatment planning and prognosis. Poorly differentiated HCC is associated with increased tumor invasiveness, early recurrence, and poor survival outcomes.\u003c/p\u003e\u003cp\u003eContrast-enhanced CT is widely used for the diagnosis and staging of HCC because of its accessibility and high spatial resolution. However, conventional CT interpretation relies primarily on qualitative image assessment and may not adequately capture intratumoral heterogeneity.\u003c/p\u003e\u003cp\u003eRadiomics refers to the high-throughput extraction of quantitative imaging features from medical images using advanced computational algorithms. These features may reflect tumor heterogeneity and biological behavior.\u003c/p\u003e\u003cp\u003eRecent studies have explored the role of radiomics in predicting tumor grade and clinical outcomes in hepatocellular carcinoma. However, the potential value of CT-based radiomics for simultaneously predicting tumor differentiation and postoperative survival remains insufficiently investigated.\u003c/p\u003e\u003cp\u003eTherefore, the aim of this study was to develop and validate a CT-based radiomics model for predicting histological grade and postoperative survival in patients with hepatocellular carcinoma.\u003c/p\u003e "},{"header":"2 Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Patient Population\u003c/h2\u003e \u003cp\u003eThis retrospective study was approved by the Institutional Review Board of Ningxia Medical University General Hospital. The requirement for informed consent was waived due to the retrospective nature of the study.\u003c/p\u003e \u003cp\u003eA total of 385 patients with pathologically confirmed hepatocellular carcinoma who underwent preoperative CT examinations between January 2013 and September 2022 were included.\u003c/p\u003e \u003cp\u003eTumor differentiation was classified according to the Edmondson\u0026ndash;Steiner grading system. Grade I\u0026ndash;II tumors were considered low-grade HCC, whereas grade III\u0026ndash;IV tumors were classified as high-grade HCC.\u003c/p\u003e \u003cp\u003ePatients were randomly assigned to a training cohort (n\u0026thinsp;=\u0026thinsp;265) and a testing cohort (n\u0026thinsp;=\u0026thinsp;120) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eA subgroup of 60 patients who underwent hepatectomy with complete follow-up data was included for survival analysis.\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\u003eClinical characteristics of patients in the training and testing cohorts.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFeature\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTraining Cohort (n\u0026thinsp;=\u0026thinsp;265)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTesting Cohort (n\u0026thinsp;=\u0026thinsp;120)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTotal (n\u0026thinsp;=\u0026thinsp;385)\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\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMale: 103, Female: 32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMale: 46, Female: 16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMale: 149, Female: 48\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAFP (ng/mL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedian (Q1, Q3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMedian (Q1, Q3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMedian (Q1, Q3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTumor Grade\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow-grade: 135, High-grade: 130\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLow-grade: 55, High-grade: 65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLow-grade: 190, High-grade: 195\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRad-score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 CT Image Acquisition\u003c/h2\u003e \u003cp\u003eAll CT examinations were performed using a 256-slice spiral CT scanner (Brilliance iCT, Philips Healthcare). Portal venous phase images were used for radiomics analysis because they provide optimal tumor-to-liver contrast for lesion delineation.\u003c/p\u003e \u003cp\u003eImages were reconstructed with a slice thickness of \u0026le;\u0026thinsp;2 mm.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Tumor Segmentation\u003c/h2\u003e \u003cp\u003eTumor segmentation was performed using ITK-SNAP software. Regions of interest were manually delineated along the tumor boundary on each CT slice by two experienced radiologists.\u003c/p\u003e \u003cp\u003eThree-dimensional volumes of interest were subsequently generated.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Radiomics Feature Extraction\u003c/h2\u003e \u003cp\u003eRadiomics features were extracted according to the Image Biomarker Standardisation Initiative (IBSI) guidelines.\u003c/p\u003e \u003cp\u003eA total of 396 radiomics features were extracted, including:\u003c/p\u003e \u003cp\u003e\u0026bull; shape features\u003c/p\u003e \u003cp\u003e\u0026bull; first-order histogram features\u003c/p\u003e \u003cp\u003e\u0026bull; texture features derived from gray-level matrices\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Feature Selection and Model Construction\u003c/h2\u003e \u003cp\u003eFeature reproducibility was assessed using ICC analysis. Features with ICC\u0026thinsp;\u0026gt;\u0026thinsp;0.75 were retained.\u003c/p\u003e \u003cp\u003eLASSO regression with ten-fold cross-validation was applied to reduce feature dimensionality and prevent overfitting.\u003c/p\u003e \u003cp\u003eSeven optimal features were selected to construct the radiomics signature (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRadiomics feature selection process.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFeature Type\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFeature Name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDescription\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShape Features\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVolume\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eVolume of the tumor (cm\u0026sup3;)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSurface Area\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSurface area of the tumor (cm\u0026sup2;)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSphericity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSphericity of the tumor\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFirst-Order Features\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean Intensity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAverage pixel intensity in ROI\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eStandard Deviation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eStandard deviation of intensity\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSkewness\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSkewness of pixel intensity\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTexture Features\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHomogeneity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHomogeneity of the tumor texture\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEntropy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eEntropy of the tumor texture\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCorrelation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCorrelation of pixel intensities\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Model Evaluation\u003c/h2\u003e \u003cp\u003eThe predictive performance of the radiomics model was evaluated using ROC analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7 Survival Analysis\u003c/h2\u003e \u003cp\u003eOverall survival was defined as the interval from surgery to death or last follow-up. Disease-free survival was defined as the interval from surgery to tumor recurrence or death.\u003c/p\u003e \u003cp\u003eKaplan\u0026ndash;Meier survival curves were generated and Cox regression analysis was used to identify independent prognostic factors (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMultivariable Cox regression analysis for overall survival.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFeature\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHazard Ratio (HR)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e95% Confidence Interval (CI)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\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\u003eDif.Scale1.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.319\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.143\u0026ndash;0.701\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDif.Scale1.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.432\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.227\u0026ndash;0.832\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.017\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDif.Scale1.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.526\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.238\u0026ndash;0.818\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.045\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eText.Homogeneity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.202\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.983\u0026ndash;1.459\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.073\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eText.Entropy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.087\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.929\u0026ndash;1.264\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.295\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShape.Volume\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.997\u0026ndash;1.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.432\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"3 Results","content":" \u003cp\u003eThe radiomics score was significantly higher in patients with high-grade HCC than in those with low-grade tumors (P\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003eThe radiomics workflow is illustrated in Fig.\u0026nbsp;1.\u003c/p\u003e \u003cp\u003e \u003cdiv description=\"figure 1\" class=\"Drawing\" id=\"1\" name=\"图片 1\"\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe categories of extracted radiomics features are shown in Fig.\u0026nbsp;2.\u003c/p\u003e \u003cp\u003e \u003cdiv description=\"figure2\" class=\"Drawing\" id=\"2\" name=\"图片 2\"\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe evaluation of the radiomics signature is presented in Fig.\u0026nbsp;3.\u003c/p\u003e \u003cp\u003e \u003cdiv description=\"figure 3\" class=\"Drawing\" id=\"3\" name=\"图片 3\"\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe model achieved an AUC of 0.889 in the training cohort and 0.941 in the testing cohort (Fig.\u0026nbsp;4).\u003c/p\u003e \u003cp\u003e \u003cdiv description=\"figure 4\" class=\"Drawing\" id=\"4\" name=\"图片 4\"\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003ePrognostic Analysis\u003c/p\u003e \u003cp\u003eAmong the 60 patients included in survival analysis, 40 experienced recurrence and 22 died during follow-up.\u003c/p\u003e \u003cp\u003eCox regression analysis identified Dif.Scale1.2 as an independent predictor of overall survival (Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eKaplan\u0026ndash;Meier curves for overall survival are shown in Fig.\u0026nbsp;5.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePrognostic performance of the radiomics model.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHazard Ratio (HR)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e95% Confidence Interval\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\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\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.94\u0026ndash;1.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.215\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTumor size\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.05\u0026ndash;2.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.024\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAFP level\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.97\u0026ndash;1.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.078\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics feature (Dif.Scale1.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.319\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.143\u0026ndash;0.701\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\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\u003e \u003cdiv description=\"figure 5\" class=\"Drawing\" id=\"5\" name=\"图片 5\"\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eKaplan\u0026ndash;Meier curves for disease-free survival are shown in Fig.\u0026nbsp;6.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"4 Discussion","content":" \u003cp\u003eThis study developed and validated a CT-based radiomics model for predicting histological grade and postoperative survival in hepatocellular carcinoma. The results demonstrate that radiomics features derived from CT images can effectively differentiate tumor grade and provide prognostic information.\u003c/p\u003e \u003cp\u003eTumor heterogeneity plays an important role in tumor progression and treatment response. Radiomics enables quantitative characterization of tumor heterogeneity from routine medical images.\u003c/p\u003e \u003cp\u003eOur findings are consistent with previous studies demonstrating the predictive value of radiomics in hepatocellular carcinoma.\u003c/p\u003e \u003cp\u003eSeveral limitations should be acknowledged. First, this was a retrospective single-center study. Second, the survival cohort was relatively small. Third, external validation was not performed.\u003c/p\u003e \u003cp\u003eFuture multicenter studies with larger sample sizes are warranted to further validate these findings. Therefore, CT-based radiomics may provide a useful imaging biomarker for individualized risk stratification in patients with hepatocellular carcinoma.\u003c/p\u003e "},{"header":"5 Conclusion","content":"\u003cp\u003eCT-based radiomics may serve as a promising noninvasive imaging biomarker for predicting tumor differentiation and postoperative survival in hepatocellular carcinoma.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eEthics approval and consent to participate\u003c/h2\u003e \u003cp\u003eThis retrospective study was approved by the Institutional Review Board of Ningxia Medical University General Hospital (Ethics approval number: KYLL-2024-0665). The requirement for informed consent was waived due to the retrospective nature of the study.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eConsent for publication\u003c/strong\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e \u003c/p\u003e\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eSun Mingxing performed the data collection, radiomics analysis, statistical analysis, and drafted the manuscript. Wang Zhijun conceived and supervised the study and critically revised the manuscript. Both authors read and approved the final manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLambin P, Rios-Velazquez E, Leijenaar R, Carvalho S, van Stiphout RGPM, Granton P, et al. Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer. 2012;48:441\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGillies RJ, Kinahan PE, Hricak H. Radiomics: images are more than pictures, they are data. Radiology. 2016;278:563\u0026ndash;77.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAerts HJWL, Velazquez ER, Leijenaar RTH, Parmar C, Grossmann P, Carvalho S, et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat Commun. 2014;5:4006.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTibshirani R. Regression shrinkage and selection via the LASSO. J R Stat Soc B. 1996;58:267\u0026ndash;88.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology. 1982;143:29\u0026ndash;36.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457\u0026ndash;81.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide. CA Cancer J Clin. 2018;68:394\u0026ndash;424.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide. CA Cancer J Clin. 2021;71:209\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLlovet JM, Kelley RK, Villanueva A, Singal AG, Pikarsky E, Roayaie S, et al. Hepatocellular carcinoma. Nat Rev Dis Primers. 2021;7:6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEl-Serag HB. Epidemiology of hepatocellular carcinoma. Gastroenterology. 2012;142:1264\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRonot M, Vilgrain V. Hepatocellular carcinoma: diagnostic criteria and imaging techniques. Radiology. 2016;279:324\u0026ndash;40.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEuropean Association for the Study of the Liver. EASL Clinical Practice Guidelines: management of hepatocellular carcinoma. J Hepatol. 2018;69:182\u0026ndash;236.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKudo M. Imaging diagnosis of hepatocellular carcinoma and premalignant/borderline lesions. Liver Cancer. 2013;2:365\u0026ndash;77.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLimkin EJ, Sun R, Dercle L, Zacharaki EI, Robert C, Reuz\u0026eacute; S, et al. Promises and challenges for the implementation of computational medical imaging (radiomics) in oncology. Nat Rev Clin Oncol. 2017;14:749\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYip SSF, Aerts HJWL. Applications and limitations of radiomics. Phys Med Biol. 2016;61:R150\u0026ndash;66.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMao B, Zhang L, Ning P, Ding F, Wu F, Lu G, et al. Preoperative prediction for pathological grade of hepatocellular carcinoma via machine learning-based radiomics. Eur Radiol. 2020;30:6924\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang YQ, Liang CH, He L, Tian J, Liang CS, Chen X, et al. Development and validation of a radiomics nomogram for preoperative prediction of microvascular invasion in hepatocellular carcinoma. Radiology. 2019;292:573\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu X, Zhang HL, Liu QP, Sun SW, Zhang J, Zhu FP, et al. Radiomic analysis of contrast-enhanced CT predicts microvascular invasion in hepatocellular carcinoma. J Hepatol. 2019;70:1133\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEdmondson HA, Steiner PE. Primary carcinoma of the liver: a study of 100 cases among 48,900 necropsies. Cancer. 1954;7:462\u0026ndash;503.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChernyak V, Fowler KJ, Kamaya A, Kielar AZ, Elsayes KM, Bashir MR, et al. LI-RADS version 2018 imaging algorithm. Radiology. 2018;289:816\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYushkevich PA, Piven J, Hazlett HC, Smith RG, Ho S, Gee JC, et al. User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. NeuroImage. 2006;31:1116\u0026ndash;28.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZwanenburg A, Valli\u0026egrave;res M, Abdalah MA, Aerts HJWL, Andrearczyk V, Apte A, et al. The image biomarker standardisation initiative. Radiology. 2020;295:328\u0026ndash;38.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Griethuysen JJM, Fedorov A, Parmar C, Hosny A, Aucoin N, Narayan V, et al. Computational radiomics system to decode the radiographic phenotype. Cancer Res. 2017;77:e104\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKumar V, Gu Y, Basu S, Berglund A, Eschrich SA, Schabath MB, et al. Radiomics: the process and the challenges. Magn Reson Imaging. 2012;30:1234\u0026ndash;48.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu M, Tan H, Gao F, Hai J, Ning P, Chen J, et al. Predicting the grade of hepatocellular carcinoma using radiomics. Front Oncol. 2020;10:177.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLubner MG, Smith AD, Sandrasegaran K, Sahani DV, Pickhardt PJ. CT texture analysis: definitions, applications, biologic correlates, and challenges. Radiographics. 2017;37:1483\u0026ndash;503.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZwanenburg A. Radiomics in nuclear medicine imaging. Eur J Nucl Med Mol Imaging. 2019;46:2638\u0026ndash;46.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParekh VS, Jacobs MA. Radiomics: a new application from established techniques. Expert Rev Precis Med Drug Dev. 2016;1:207\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun R, Limkin EJ, Vakalopoulou M, Dercle L, Champiat S, Han SR, et al. A radiomics approach to assess tumour heterogeneity. Nat Rev Clin Oncol. 2020;17:757\u0026ndash;72.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParmar C, Grossmann P, Bussink J, Lambin P, Aerts HJWL. Machine learning methods for quantitative radiomic biomarkers. Nat Rev Clin Oncol. 2015;12:104\u0026ndash;17.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Radiomics, Computed tomography, Hepatocellular carcinoma, Tumor differentiation, Prognosis, Survival prediction","lastPublishedDoi":"10.21203/rs.3.rs-9078107/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9078107/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eTo develop and validate a computed tomography (CT)-based radiomics model for predicting histological grade and postoperative survival in patients with hepatocellular carcinoma (HCC).\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThis retrospective study included 385 patients with pathologically confirmed HCC who underwent preoperative CT examinations between January 2013 and September 2022. Patients were randomly assigned to a training cohort (n\u0026thinsp;=\u0026thinsp;265) and a testing cohort (n\u0026thinsp;=\u0026thinsp;120). Radiomics features were extracted from portal venous phase CT images using standardized radiomics pipelines. Feature selection was performed using intraclass correlation coefficient (ICC) analysis and least absolute shrinkage and selection operator (LASSO) regression. Model performance was evaluated using receiver operating characteristic (ROC) analysis. Survival outcomes were assessed using Kaplan\u0026ndash;Meier analysis and Cox proportional hazards regression.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eAmong 396 extracted radiomics features, 180 features demonstrated good reproducibility (ICC\u0026thinsp;\u0026gt;\u0026thinsp;0.75). After dimensionality reduction, seven radiomics features were selected to construct the radiomics signature. The model achieved an AUC of 0.889 in the training cohort and 0.941 in the testing cohort. Survival analysis identified the radiomics feature Dif.Scale1.2 as an independent predictor of overall survival (HR\u0026thinsp;=\u0026thinsp;0.319, 95% CI: 0.143\u0026ndash;0.701, P\u0026thinsp;=\u0026thinsp;0.002).\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eCT-based radiomics may serve as a promising noninvasive imaging biomarker for predicting tumor differentiation and postoperative survival in hepatocellular carcinoma.\u003c/p\u003e","manuscriptTitle":"CT-Based Radiomics Model for Preoperative Prediction of Histological Grade and Postoperative Survival in Hepatocellular Carcinoma","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-25 08:20:43","doi":"10.21203/rs.3.rs-9078107/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":"eff12977-d719-45e7-9b13-7488a54bbced","owner":[],"postedDate":"March 25th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Rejected","date":"2026-05-08T06:11:44+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-08T06:24:47+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-25 08:20:43","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9078107","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9078107","identity":"rs-9078107","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}