CT study of radiomics features predicting Ki-67 expression level in peripheral lung cancer

CT study of radiomics features predicting Ki-67 expression level in peripheral lung cancer | 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 study of radiomics features predicting Ki-67 expression level in peripheral lung cancer Lihua Fan, Wei Wei, Jian Geng, Dong Han, Yongjun Jia, Nan Yu, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4323989/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 Background The association between radiomics features of peripheral lung cancer and Ki-67 expression level remains unclear. To develop a radiomics signature based on enhanced CT arterial phase image to estimate the expression situation of Ki-67 in peripheral lung cancer. Methods A total of 117 peripheral lung cancer patients, who underwent contrast-enhanced CT scan in our hospital from May 2016 to November 2019, including 43 males and 74 females, aged 35 to 79 years old (median 54 years old). All the peripheral lung cancers were confirmed by histopathological and took in Ki-67 expression situation detection within 2 weeks after CT inspection, including 63 cases of Ki-67 low expression and 54 cases of Ki-67 high expression, which were retrospectively analyzed and were divided into training ( n = 82) and validation cohorts ( n = 35) in a ratio of 7:3. ITK-SNAP was used to manually outline the total tumor volume data of lung cancer on CT arterial phase images, and the radiomics features were extracted by A.K software. LASSO regression model was used to further screen features and construct radiomics labels, and the radiomics score of each patient was calculated, and then multi-factor logistic regression analysis was performed combined with clinical information to screen out independent risk factors for predicting Ki-67 levels. The predictive accuracy of the radiomics signature was quantified by the area under the curve (AUC) of receiver operator characteristic (ROC) curve in both the training and validation cohorts. The Hosmer-Lemeshow test was performed to evaluate the calibration degree of the radiomics. We performed decision curve analysis (DCA) to assess the clinical usefulness of the radiomics signature. Results Seven radiomics features were chosen from 396 candidate features to build a radiomics label that significantly correlated with Ki-67 expression level. The model showed good calibration and discrimination in the training cohort, with an AUC of 0.844 (95%CI: 0.725–0.964), sensitivity of 93% and specificity of 71%, calibration degree of 0.709. In the validation cohort, AUC was 0.881 (95%CI: 0.756–0.954), sensitivity was 91%, and specificity was 75%, calibration degree of 0.950. Univariate logistic regression analysis showed that there were no conspicuous differences in gender, age and smoke between the high and low Ki-67 expression ( P > 0.05). Using multivariate logistic regression model, radiomics signature were considered to be independent predictor of Ki-67 expression level in peripheral lung cancer. DCA for the radiomics signature in the training cohort showed that if the threshold probability was between 0.03 and 0.63, then using the radiomics signature to predict Ki-67 expression situation added more benefit than treating either all or no patients. Conclusion The radiomics signature based on enhanced CT arterial phase image is benefitial to predict the expression of Ki-67 in peripheral lung cancer, which can assess the invasiveness and prognosis for peripheral lung cancer noninvasively. Radiomics Ki-67 proliferation index Peripheral lung cancer Tomography X-ray computed Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction Lung cancer is the leading cause of cancer death worldwide[ 1 ], among which peripheral lung cancer accounts for more than 70%. The early symptoms of peripheral lung cancer are not significant, and the first diagnosis is often advanced, so early diagnosis and timely treatment are very important. Although there are more and more treatment methods for lung cancer, some patients with early stage lung cancer still have poor prognosis[ 2 ], and the most important reason may be the heterogeneity of cancer in space and time[ 3 ]. With the clinical application of precision therapy, molecular pathology has been used as the gold standard to replace the traditional surgical model[ 4 ]. Many representative biological markers are produced when tumor cells proliferate, differentiate and spread[ 5 ]. Therefore, more convenient molecular biological markers are needed to dynamically observe the biological behavior of lung cancer. Uncontrolled cell proliferation is the most direct cause of cancer occurrence and development. Ki-67 is the most commonly used marker to reflect the proliferative state of lung cancer lesions[ 6 ], evaluate prognosis and design molecular targeted drugs. However, the expression level of Ki-67 can only be determined by postoperative tissue samples, which may lead to bleeding, pneumothorax and increase the possibility of metastasis[ 7 ]. Therefore, it is very important to find a non-invasive method to predict the expression level of Ki-67 before surgery. CT enhanced imaging is the most commonly used method in the diagnosis of lung cancer[ 8 ]. It has been confirmed in the literature that with the increase of Ki-67 expression level, the enhancement degree of CT is higher, the differentiation degree is lower, and the enhancement is more uneven[ 9 ]. Compared with traditional images, radiomics can better display the internal characteristics of the lesion and provide a large amount of information about the shape, size, strength and volume of the tumor[ 10 ], which can accurately evaluate the biological behavior of the tumor and cancer genetics[ 11 ], reflecting the advantages of non-invasive and convenient. Therefore, this study was mainly based on CT enhanced arterial phase images to construct imaging labels to predict Ki-67 expression levels in peripheral lung cancer, providing more evidence for evaluating tumor invasiveness and prognosis. Methods Materials This study was approved by our Institutional Ethics Committee (Approval number: SZFYIEC-YJ-KYBC-2022 No.07), and written informed consent was obtained from all participants. A total of 117 patients with peripheral lung cancer who received chest CT plain scan and enhanced scan from May 2017 to November 2020 and were pathologically confirmed and Ki-67 expression level detected within 2 weeks after examination were retrospectively collected, and met the following inclusion and exclusion criteria. Inclusion criteria: 1) peripheral lung cancer patients with complete enhanced CT arterial phase image data, Ki-67 expression results and pathological results confirmed by surgery or puncture biopsy; 2) The lesions were solid nodules and masses; 3) Did not receive radiotherapy or chemotherapy treatment. Exclusion criteria: 1) combined with primary malignant tumor of other organs; 2) The image quality was poor due to respiratory movement artifacts, which affected lesion segmentation; 3) The tumor was invasive and difficult to segment. Finally, a total of 117 patients with peripheral lung cancer were included, including 43 males and 74 females, aged 35 to 79 years old (median 54 years old). Pathology confirmed that 54 patients in the Ki-67 high expression group and 63 patients in the Ki-67 low expression group were randomly divided into the training cohort (n = 82) and the validation cohort (n = 35) by stratified sampling method at a ratio of 7:3. Acquisition of image data Informed consent for enhanced CT was signed by all patients before CT scan. GE Discovery CT 750 HD (GE Healthcare, OH) was used to perform routine plain scan and two-phase enhanced scan after 30 s and 60 s injection of contrast agent. Scan from thoracic entrance to lowest lateral costophrenic Angle level. Scanning parameters: layer thickness of 5 mm, layer spacing of 5 mm; The reconstruction thickness was 1 mm, the reconstruction distance was 1 mm, the tube voltage was 120 kV, and the tube current was automatic mA. The non-ionic iodinated contrast agent Ioversol (350 mg I/ml; Jiangsu Hengrui Medicine Co. Ltd, China) in the dosage of 1.14 ml/kg was used, which was injected through the right elbow vein with a dual-tube power injector (Ulrich, Germany) at an injection rate of 3.5-4.0 ml/s. Ki-67 immunohistochemical analysis The pathological results of all patients were confirmed by puncture or postoperative pathology, and 3 sites were selected to ensure that Ki-67 detection was representative of the entire tumor. Ki-67 expression level was divided into high expression group (Ki-67 ≥ 50%) and low expression group (Ki-67 < 50%). Delineation of region of interest and calculation of radiomics features The CT enhanced arterial phase images were imported into ITK-Snap software, and then the volume of interest (VOI) of the tumor was mapped by a radiologist with 8 years of experience in chest imaging diagnosis, including lobular, cavity, hemorrhage, necrosis, and vascular areas. The VOI of the tumor was saved as a boundary file (.nii file format), which was manually modified by another intermediate or higher diagnosticist after two months to ensure the accuracy of the boundary file. The DICOM data and boundary files of tumor images were imported into A.K software, and 396 radiomics features were extracted, including 6 types of features (Fig. 1 ), in which GLCM and GLRLM steps were set to 1, 4 and 7. Radiomics feature screening and model building In this study, LASSO regression model was used to further screen the radiomics features of peripheral lung cancer lesions in the training cohort, and the radiomics signature was established based on the screening radiomics features and their regression coefficients, and the imaging scores of each patient were calculated, and binary and multi-factor Logistic regression analysis was performed combined with basic clinical data. Independent risk factors were selected to predict Ki-67 expression levels. The radiomics model was used to internally validate the data of the validation cohort. Evaluation of radiomics signature performance The sensitivity and specificity were calculated according to the receiver operating characteristic (ROC) curve, and the predictive ability of radiomics signature was observed using the area under curve (AUC). The calibrations of the radiomics signature in the training and validation cohorts were evaluated according to the Hosmer-Lemeshow test. Decision curve analysis (DCA)[ 12 ] was used to analyze the net benefit of patients under different probability thresholds, so as to evaluate the clinical value of radiomics signature. Statistical analysis SPSS 22.0 software and R language were used for statistical analysis, and P < 0.05 was considered statistically significant. The measurement data satisfying normal distribution and homogeneity of variance were expressed as mean ± standard deviation using independent sample t test. Counting data were measured by chi-square test and expressed as n /%. For consistency test, binary classification parameters were tested by Kappa test, and ordinal multi-classification parameters were tested by Weighted-Kappa test. Results Risk factors for predicting Ki-67 expression in peripheral lung cancer Ki-67 expression level was divided into high expression group and low expression group in training cohort and verification group. Logistic regression analysis of patient in gender, age, smoking and radiomics scores showed that there were no significant differences in age, gender and smoking between the training cohort and the verification group ( P >0.05). Radiomics scores were considered to be independent predictors of Ki-67 expression levels in peripheral lung cancer (Table 1). Table 1 Prediction of risk factors for Ki-67 in peripheral lung cancer Variables and intercepts One-factor Logistic regression Multi-factor Logistic regression OR（95% CI） p value OR（95% CI） p value Gender 0.693（0.135-3.553） 0.660 NA NA Age 0.982（0.899-1.073） 0.687 NA NA smoking and 0.798（0.534-5.361） 0.583 NA NA Radiomics scores 3.935（4.469-53.495） <0.01 3.935（4.469-53.495） <0.001 Note：OR：odds ratio；NA：not available. Radiomics scores The Z-scores of 396 radiomics features were normalized, the dimensionality of the data was reduced by LASSO regression model, and the optimal λ value was obtained by cross-validation (Fig. 2). A total of 7 non-zero coefficient features were screened (Fig. 3). The CT arterial phase images and radiomics features of Ki-67 high and low expression groups were shown.(Figure. 4 and Figure. 5). Then the radiomics risk score was calculated using the radiomics signature. The radiomics scores of the Ki-67 high and low expression groups were -1.89±1.13 and 0.51±1.18 in the training cohort, with statistical differences ( t =-7.89, P <0.01), and -1.28±1.21 and 0.74±0.87 in the validation cohort, with statistical differences ( t =-3.76, P <0.01). Radiomics prediction model formula Ki-67 expression level was taken as the dependent variable, and the radiomics score of peripheral lung cancer was taken as the independent variable to be included in the binary Logistics regression. An independent predictor of Ki-67 expression in peripheral lung cancer was the imaging score. The prediction model was Y=1/ [1+exp (-Z)], Z=2.275 * Radiomics_score + 0.712. Radiomics signature predict performance ROC analysis of radiomics signature showed that AUC in the training cohort was 0.844 (95%CI: 0.725-0.964), sensitivity was 93%, and specificity was 71%. The Hosmer-Lemeshow test results showed that the calibration was 0.709 (Fig. 6). In the validation cohort, AUC was 0.881 (95%CI: 0.756-0.954), sensitivity was 91%, specificity was 75%, and calibration was 0.950 (Fig. 7). In both cohorts, the radiomics signature could predict the level of Ki-67 expression well. Decision curve analysis The analysis of DCA images showed that when the threshold probability was between 3% and 63%, radiomics signature based on CT enhanced arterial phase images predicted Ki-67 expression in peripheral lung cancer better than all patients considered Ki-67 high expression and all patients considered Ki-67 low expression (Fig. 8). Discussion With the increasing incidence of peripheral lung cancer, early screening, prediction and prognosis of this disease are particularly important[ 13 ]. Therefore, when peripheral lung cancer is found, it has certain value for subsequent treatment to judge its malignancy degree and prognosis by using certain radiography examination and biological markers. Although the CT findings of peripheral lung cancer can detect nodules or masses in the chest, due to the relatively complex image features, it is difficult to identify benign and malignant, so it is necessary to diagnose peripheral lung cancer by radiography, tumor biological markers, pathological tissue section staining and other methods[ 14 ]. Ki-67, as a biomarker related to cell proliferation[ 15 ], is one of the most studied markers of proliferating cells. Studies[ 16 ] had shown that Ki-67 proliferation index was significantly correlated with the degree of differentiation and invasiveness of tumors. The higher the expression level of Ki-67, the faster the tumor growth, the higher the malignancy, the stronger the invasiveness, and the worse the prognosis, so it could be used to assess the degree of malignancy of the tumor. Ki-67 also plays an important role in the selection of tumor treatment and prognosis evaluation. Because of its convenient and non-invasive advantages, enhanced CT has become a common radiography examination method in clinical diagnosis of lung cancer. Different lung cancers have different growth patterns, so their CT findings show different imaging features. The expression of Ki-67 and other tumor markers determines the tumor growth pattern. Since the concept of "radiomics" was first proposed by Gillies et al.[ 17 ] in 2010 and subsequently improved by Lambin et al.[ 18 ], it had gradually developed into a research hotspot in the field of radiography. By analyzing images using radiomics method, molecular level information contained in images could be obtained. It could assist clinicians to make independent treatment plans for patients[ 19 ]. Chest enhanced CT images were analyzed by radiomics, and different image features had significant effect in predicting the expression level of molecular biological marker Ki-67, and could guide clinical treatment and prognosis. In this study, radiomics signature based on enhanced CT arterial phase images were used to predict the expression level of Ki-67 in peripheral lung cancer. The results of this study were as follows: The ROC curve of radiomics signature in predicting Ki-67 expression level of peripheral lung cancer showed that the AUC in the training cohort was 0.844, and the AUC in the validation cohort was 0.881. In the two test cohorts, radiomics signature could well predict the level of Ki-67 expression in peripheral lung cancer, and then predict the malignancy and prognosis of the tumor. It provided some guiding value for clinicians to formulate the next treatment measures. In this study, A.K software was used to extract lung cancer radiomics features, and then LASSO regression was used to further reduce the dimension of the feature datas and establish the model. Certain penalty terms were added to the cost function to make the established model have a sharper function, so it was not easy to produce certain overfitting. The optimal λ value was obtained by cross-validation, and the radiomics prediction model was closer to the ideal model. Previous studies had shown that radiomics score was an independent risk factor for predicting malignant tumors[ 20 ]. In this study, only radiomics score was an independent risk factor for predicting Ki-67 expression based on the radiomics prediction model, which was consistent with previous studies. Some studies had shown that the expression level of Ki-67 was closely related to clinical data (gender, age, and smoking)[ 21 ], but in this study, there was no statistical significance in age, gender, and smoking between the groups with high or low expression of Ki-67 ( P > 0.05), which contradicted the results of previous studies and may be related to the small number of cases in this study. Some studies had found that Ki-67 was an independent risk factor for recurrence in patients with early lung adenocarcinoma after segmentation resection[ 22 ]. Therefore, it is very important for the selection of surgical methods to determine the expression of Ki-67 by non-invasive detection before surgery. In this study, we found that the radiomics signature can predict the expression level of Ki-67 noninvasively, which can guide clinicians to choose the appropriate surgical method. This study has the following limitations: First, the number of cases is small, so there is bias in this study. Second, this study takes peripheral lung cancer as the research object and has not included specific pathological types. Third, this study only used enhanced CT arterial phase images for feature extraction, and subsequently include other phases in the study. Fourth, some cases in this study were confirmed by percutaneous puncture biopsy, and although it can also be used to predict Ki-67 expression level, small biopsy tissues cannot represent the tumor tissue itself, so it is still challenging to predict Ki-67 expression level. Therefore, in the follow-up study, increasing the sample size, adding specific pathological types and detailed clinical and image-related information will help improve the detection performance of radiomics signature. In summary, this study showed that it has a high diagnostic value in predicting the Ki-67 expression level of peripheral lung cancer based on the radiomics signature created by enhanced CT arterial phase images, and provides a non-invasive examination means to assist clinicians in making a dedicated treatment plan for patients and evaluating prognosis. Abbreviations AUC area under the curve ROC receiver operator characteristic DCA decision curve analysis VOI volume of interest GLCM gray-level co-occurrence matrix GLRLM gray-level run length matrix GLSZM gray-level size zone matrix Declarations Acknowledgements We sincerely appreciate all the patients who participated in this study. Funding This work was supported by the Youth Innovation Team Scientific Research project of Shaanxi Provincial Education Department (grant no. 23JP035). Availability of data and materials Data and materials generated or analyzed during the study are available from the corresponding author by request. Ethics approval and consent to participate The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the ethics committee of Affiliated Hospital of Shaanxi University of Chinese Medicine, and written informed consent was obtained from all participants. Competing interests The authors declare no competing interests. Consent to publish Not applicable. Conflict of interest The authors do not have any conflicts of interest. Authors’ contributions All authors contributed to the study's conception and design. Material preparation, data collection, and analysis were performed by Wei Wei, Jian Geng, Dong Han, Yongjun Jia, Nan Yu, Shan Dang, Yong Yu, Yunsong Zheng and Lihua Fan. The first draft of the manuscript was written by Wei Wei, Lihua Fan and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. References Thai AA, Solomon BJ, Sequist LV, Gainor JF, Heist RS. Lung cancer. Lancet. 2021;398:535–54. K W JR. Y X, M Z, T R, Z G, Simulation framework for generating intratumor heterogeneity patterns in a cancer cell population. Eur J Surg oncology: J Eur Soc Surg Oncol Br Association Surg Oncol. 2023;49. Wm I. H I. Simulation framework for generating intratumor heterogeneity patterns in a cancer cell population. PLoS ONE. 2017;12. T KA, Ma RAVIPPC. B, Circulating Tumour Cells: Detection and Application in Advanced Non-Small Cell Lung Cancer. Int J Mol Sci. 2023;24. Eec de SW, Je J, van A I T. I C, J P, Decision Support Systems in Oncology. JCO Clin cancer Inf. 2019;3. Jkt D, Cf F. The Role of Histologic Grading and Ki-67 Index in Predicting Outcomes in Pulmonary Carcinoid Tumors. Am J Surg Pathol. 2020;44. Kobara H, Mori H, Rafiq K, Fujihara S, Nishiyama N, Chiyo T, et al. Analysis of the amount of tissue sample necessary for mitotic count and Ki-67 index in gastrointestinal stromal tumor sampling. Oncol Rep. 2015;33:215–22. Zp M, Xl L, Tl KG, Hd Z, Yx W. Z. Application of radiomics based on chest CT-enhanced dual-phase imaging in the immunotherapy of non-small cell lung cancer. J X-Ray Sci Technol. 2023;31. D U-G AA. N B, M D. Expression of Ki-67 and Estrogen Receptor Beta in Primary Cutaneous Melanoma as a Potential Indicator of Regional Lymph Node Positivity. Applied immunohistochemistry & molecular morphology: AIMM. 2019;27. Ma M, Mr AS, Eh H, Pk C Jr. M. Association of distant recurrence-free survival with algorithmically extracted MRI characteristics in breast cancer. J Magn Reson imaging: JMRI. 2019;49. Nm B, C MEPP. D, H G, P T, Intratumoral and peritumoral radiomics for the pretreatment prediction of pathological complete response to neoadjuvant chemotherapy based on breast DCE-MRI. Breast cancer research: BCR. 2017;19. Br L, Mk S, Br H, M Z, Ac N, Sc C. Renal mass biopsy–a renaissance? J Urol. 2008;179. Mc AM Jr. M, H R, G L, K S, Probability of cancer in pulmonary nodules detected on first screening CT. N Engl J Med. 2013;369. C P, Cm RG et al. G, A V, M P, F G,. CT-guided transthoracic needle biopsy: advantages in histopathological and molecular tests. Future oncology (London, England). 2020;16. Mz M, Ah N, N N. MCM – 2 and Ki – 67 as proliferation markers in renal cell carcinoma: A quantitative and semi - quantitative analysis. Int Braz J Urol. 2016;42. Rg B, Gm C, Ac G, Ct C, An C, Am A et al. Histologic correlation of expression of Ki-67 in squamous cell carcinoma of the glottis according to the degree of cell differentiation. Braz J Otorhinolaryngol. 2014;80. Rj G, Pe K. H H. Radiomics: Images Are More than Pictures, They Are Data. Radiology. 2016;278. E R-V PL, Rg van RLSC et al. S, P G,. Radiomics: extracting more information from medical images using advanced feature analysis. European journal of cancer (Oxford, England: 1990). 2012;48. Q MZ, L Y CYG. Y T, C L, Development and validation of CT-based radiomics nomogram for the classification of benign parotid gland tumors. Med Phys. 2023;50. H X, Y QL. Z, J H, Y S, M L, Noninvasive evaluation of neutrophil extracellular traps signature predicts clinical outcomes and immunotherapy response in hepatocellular carcinoma. Front Immunol. 2023;14. Q F, Sl L, Dp H, Yb H, Xj L, Z Z et al. CT Radiomics Model for Predicting the Ki-67 Index of Lung Cancer: An Exploratory Study. Front Oncol. 2021;11. S Y TM, K T MM. M C, S Y, Ki-67 labeling index is associated with recurrence after segmentectomy under video-assisted thoracoscopic surgery in stage I non-small cell lung cancer. Annals Thorac Cardiovasc surgery: official J Association Thorac Cardiovasc Surg Asia. 2011;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-4323989","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":299273981,"identity":"80bcfae0-6bde-46f9-bbb2-24d304d775d9","order_by":0,"name":"Lihua Fan","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0klEQVRIiWNgGAWjYBACNoaDjQ8+VEjIsbE3HyBOCx/j4WbDGWcsjPl4jiUQp0WO+XibMG9bReI8iRwDIh3GdrCNgeeMRGIbz5mPN94w2MnpNhDSwnOw7YFEhYRxG3vvZss5DMnGZgcIaZE42G5gcEZCto3n7DZpHoYDidsIapF/2AZ0lQRjm0TOMyK1MBxskwAiRaAWNqK1NBs2nJEwZuM5Zmw5x4AIv8g3HH/4+E9FnZx8e/PDG28q7OQIakEBEjxERg2yFlJ1jIJRMApGwYgAAG4LRIRD4zNFAAAAAElFTkSuQmCC","orcid":"","institution":"Affiliated Hospital of Shaanxi University of Chinese Medicine","correspondingAuthor":true,"prefix":"","firstName":"Lihua","middleName":"","lastName":"Fan","suffix":""},{"id":299273982,"identity":"5d915a60-14c8-484f-8435-0e0cf2d22e95","order_by":1,"name":"Wei Wei","email":"","orcid":"","institution":"Affiliated Hospital of Shaanxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Wei","suffix":""},{"id":299273983,"identity":"8e2a7488-2e7e-455e-96f4-fdcbff956404","order_by":2,"name":"Jian Geng","email":"","orcid":"","institution":"Affiliated Hospital of Shaanxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Jian","middleName":"","lastName":"Geng","suffix":""},{"id":299273984,"identity":"3873d66e-a561-425f-81cf-a6b6a1486ace","order_by":3,"name":"Dong Han","email":"","orcid":"","institution":"Affiliated Hospital of Shaanxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Dong","middleName":"","lastName":"Han","suffix":""},{"id":299273985,"identity":"16b8cc84-0379-4cbc-9727-14b390fb753a","order_by":4,"name":"Yongjun Jia","email":"","orcid":"","institution":"Affiliated Hospital of Shaanxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yongjun","middleName":"","lastName":"Jia","suffix":""},{"id":299273986,"identity":"c50b5f42-98ed-4f32-a7d7-a01f9134ecb1","order_by":5,"name":"Nan Yu","email":"","orcid":"","institution":"Shaanxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Nan","middleName":"","lastName":"Yu","suffix":""},{"id":299273987,"identity":"4c36702b-5264-4ed9-b605-05c51f07b13c","order_by":6,"name":"Shan Dang","email":"","orcid":"","institution":"Affiliated Hospital of Shaanxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Shan","middleName":"","lastName":"Dang","suffix":""},{"id":299273990,"identity":"f7730245-ba59-46b9-b4b9-8df609afd245","order_by":7,"name":"Yong Yu","email":"","orcid":"","institution":"Affiliated Hospital of Shaanxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yong","middleName":"","lastName":"Yu","suffix":""},{"id":299273991,"identity":"3e379eb5-3178-4a52-8dae-917860a2749c","order_by":8,"name":"Yunsong Zheng","email":"","orcid":"","institution":"Affiliated Hospital of Shaanxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yunsong","middleName":"","lastName":"Zheng","suffix":""}],"badges":[],"createdAt":"2024-04-25 12:01:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4323989/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4323989/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":56140112,"identity":"8813b872-a492-4fac-8799-8e1ef0e37591","added_by":"auto","created_at":"2024-05-09 04:03:47","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":107433,"visible":true,"origin":"","legend":"\u003cp\u003eLesion segmentation and radiomics feature extraction. A) Original DICOM images of enhanced CT arterial phase; B), C) Lesion segmentation and 3D View; D) Extract radiomics features, including gray-level co-occurrence matrix (GLCM), Histogram, gray-level run length matrix (GLRLM), gray-level size zone matrix (GLSZM), shape (Formfactor) and Haralick.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-4323989/v1/1a4eaafb8e0f756c2d52bd70.png"},{"id":56140254,"identity":"d817d114-0e43-4114-b423-462583b0d20a","added_by":"auto","created_at":"2024-05-09 04:10:03","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":710911,"visible":true,"origin":"","legend":"\u003cp\u003eCross validation graph. Log（λ）=-2.045, the optimal λ value was 0.1309.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-4323989/v1/dc3d1086a00073c47d7cfa7f.png"},{"id":56140217,"identity":"5efbe0f5-d516-4b87-8c3c-998bda15aa56","added_by":"auto","created_at":"2024-05-09 04:09:35","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":717571,"visible":true,"origin":"","legend":"\u003cp\u003eCoefficient features. A total of 7 non-zero coefficient features were screened by Log (λ).\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-4323989/v1/d3c0992c54f2b6fa7080ac9d.png"},{"id":56140174,"identity":"0b8afac1-4e57-412d-8c66-4266b715f0bb","added_by":"auto","created_at":"2024-05-09 04:08:12","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":531289,"visible":true,"origin":"","legend":"\u003cp\u003ePeripheral lung cancer in inferior lobe of right lung. The transection images of CT arterial phase and radiomics features in Ki-67 high expression group.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-4323989/v1/3b51672f85ec9b44bfafc3a5.png"},{"id":56140182,"identity":"ca2c2ddb-b5c4-4a93-a0fc-055dd6c19ca6","added_by":"auto","created_at":"2024-05-09 04:08:52","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":495772,"visible":true,"origin":"","legend":"\u003cp\u003ePeripheral lung cancer in inferior lobe of left lung. The transection images of CT arterial phase and radiomics features in Ki-67 low expression group.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-4323989/v1/ebfbc0c0525eab20c74ff694.png"},{"id":56140242,"identity":"cf45ef62-cc34-4d5b-a065-dc2919eead3e","added_by":"auto","created_at":"2024-05-09 04:09:57","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":209781,"visible":true,"origin":"","legend":"\u003cp\u003eTraining cohort\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-4323989/v1/ea457f71ceb71be871b06030.png"},{"id":56140179,"identity":"98076b18-094f-4d33-9548-721d830b181f","added_by":"auto","created_at":"2024-05-09 04:08:37","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":178748,"visible":true,"origin":"","legend":"\u003cp\u003evalidation cohort\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-4323989/v1/3318fc7a9f0b9a505dd019b3.png"},{"id":56140098,"identity":"d3050c7b-4027-451b-8528-619fcf6c1268","added_by":"auto","created_at":"2024-05-09 04:03:05","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":234685,"visible":true,"origin":"","legend":"\u003cp\u003eDecision curve\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-4323989/v1/417107133a47e81d1985fed2.png"},{"id":66537119,"identity":"e106cdde-9374-48d8-a55a-35ae844d9637","added_by":"auto","created_at":"2024-10-14 07:09:46","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5163126,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4323989/v1/a7c12866-8452-45af-b786-16d283fed529.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"CT study of radiomics features predicting Ki-67 expression level in peripheral lung cancer","fulltext":[{"header":"Introduction","content":"\u003cp\u003eLung cancer is the leading cause of cancer death worldwide[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e], among which peripheral lung cancer accounts for more than 70%. The early symptoms of peripheral lung cancer are not significant, and the first diagnosis is often advanced, so early diagnosis and timely treatment are very important. Although there are more and more treatment methods for lung cancer, some patients with early stage lung cancer still have poor prognosis[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e], and the most important reason may be the heterogeneity of cancer in space and time[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. With the clinical application of precision therapy, molecular pathology has been used as the gold standard to replace the traditional surgical model[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Many representative biological markers are produced when tumor cells proliferate, differentiate and spread[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Therefore, more convenient molecular biological markers are needed to dynamically observe the biological behavior of lung cancer.\u003c/p\u003e \u003cp\u003eUncontrolled cell proliferation is the most direct cause of cancer occurrence and development. Ki-67 is the most commonly used marker to reflect the proliferative state of lung cancer lesions[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e], evaluate prognosis and design molecular targeted drugs. However, the expression level of Ki-67 can only be determined by postoperative tissue samples, which may lead to bleeding, pneumothorax and increase the possibility of metastasis[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Therefore, it is very important to find a non-invasive method to predict the expression level of Ki-67 before surgery. CT enhanced imaging is the most commonly used method in the diagnosis of lung cancer[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. It has been confirmed in the literature that with the increase of Ki-67 expression level, the enhancement degree of CT is higher, the differentiation degree is lower, and the enhancement is more uneven[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Compared with traditional images, radiomics can better display the internal characteristics of the lesion and provide a large amount of information about the shape, size, strength and volume of the tumor[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], which can accurately evaluate the biological behavior of the tumor and cancer genetics[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], reflecting the advantages of non-invasive and convenient.\u003c/p\u003e \u003cp\u003eTherefore, this study was mainly based on CT enhanced arterial phase images to construct imaging labels to predict Ki-67 expression levels in peripheral lung cancer, providing more evidence for evaluating tumor invasiveness and prognosis.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eMaterials\u003c/h2\u003e \u003cp\u003e This study was approved by our Institutional Ethics Committee (Approval number: SZFYIEC-YJ-KYBC-2022 No.07), and written informed consent was obtained from all participants. A total of 117 patients with peripheral lung cancer who received chest CT plain scan and enhanced scan from May 2017 to November 2020 and were pathologically confirmed and Ki-67 expression level detected within 2 weeks after examination were retrospectively collected, and met the following inclusion and exclusion criteria. Inclusion criteria: 1) peripheral lung cancer patients with complete enhanced CT arterial phase image data, Ki-67 expression results and pathological results confirmed by surgery or puncture biopsy; 2) The lesions were solid nodules and masses; 3) Did not receive radiotherapy or chemotherapy treatment. Exclusion criteria: 1) combined with primary malignant tumor of other organs; 2) The image quality was poor due to respiratory movement artifacts, which affected lesion segmentation; 3) The tumor was invasive and difficult to segment. Finally, a total of 117 patients with peripheral lung cancer were included, including 43 males and 74 females, aged 35 to 79 years old (median 54 years old). Pathology confirmed that 54 patients in the Ki-67 high expression group and 63 patients in the Ki-67 low expression group were randomly divided into the training cohort (n\u0026thinsp;=\u0026thinsp;82) and the validation cohort (n\u0026thinsp;=\u0026thinsp;35) by stratified sampling method at a ratio of 7:3.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eAcquisition of image data\u003c/h2\u003e \u003cp\u003e \u003cstrong\u003eInformed consent\u003c/strong\u003e \u003cp\u003efor enhanced CT was signed by all patients before CT scan. GE Discovery CT 750 HD (GE Healthcare, OH) was used to perform routine plain scan and two-phase enhanced scan after 30 s and 60 s injection of contrast agent. Scan from thoracic entrance to lowest lateral costophrenic Angle level. Scanning parameters: layer thickness of 5 mm, layer spacing of 5 mm; The reconstruction thickness was 1 mm, the reconstruction distance was 1 mm, the tube voltage was 120 kV, and the tube current was automatic mA. The non-ionic iodinated contrast agent Ioversol (350 mg I/ml; Jiangsu Hengrui Medicine Co. Ltd, China) in the dosage of 1.14 ml/kg was used, which was injected through the right elbow vein with a dual-tube power injector (Ulrich, Germany) at an injection rate of 3.5-4.0 ml/s.\u003c/p\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eKi-67 immunohistochemical analysis\u003c/h2\u003e \u003cp\u003eThe pathological results of all patients were confirmed by puncture or postoperative pathology, and 3 sites were selected to ensure that Ki-67 detection was representative of the entire tumor. Ki-67 expression level was divided into high expression group (Ki-67\u0026thinsp;\u0026ge;\u0026thinsp;50%) and low expression group (Ki-67\u0026thinsp;\u0026lt;\u0026thinsp;50%).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eDelineation of region of interest and calculation of radiomics features\u003c/h2\u003e \u003cp\u003eThe CT enhanced arterial phase images were imported into ITK-Snap software, and then the volume of interest (VOI) of the tumor was mapped by a radiologist with 8 years of experience in chest imaging diagnosis, including lobular, cavity, hemorrhage, necrosis, and vascular areas. The VOI of the tumor was saved as a boundary file (.nii file format), which was manually modified by another intermediate or higher diagnosticist after two months to ensure the accuracy of the boundary file. The DICOM data and boundary files of tumor images were imported into A.K software, and 396 radiomics features were extracted, including 6 types of features (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003e), in which GLCM and GLRLM steps were set to 1, 4 and 7.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eRadiomics feature screening and model building\u003c/h2\u003e \u003cp\u003eIn this study, LASSO regression model was used to further screen the radiomics features of peripheral lung cancer lesions in the training cohort, and the radiomics signature was established based on the screening radiomics features and their regression coefficients, and the imaging scores of each patient were calculated, and binary and multi-factor Logistic regression analysis was performed combined with basic clinical data. Independent risk factors were selected to predict Ki-67 expression levels. The radiomics model was used to internally validate the data of the validation cohort.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eEvaluation of radiomics signature performance\u003c/h2\u003e \u003cp\u003eThe sensitivity and specificity were calculated according to the receiver operating characteristic (ROC) curve, and the predictive ability of radiomics signature was observed using the area under curve (AUC). The calibrations of the radiomics signature in the training and validation cohorts were evaluated according to the Hosmer-Lemeshow test. Decision curve analysis (DCA)[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] was used to analyze the net benefit of patients under different probability thresholds, so as to evaluate the clinical value of radiomics signature.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eSPSS 22.0 software and R language were used for statistical analysis, and \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant. The measurement data satisfying normal distribution and homogeneity of variance were expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation using independent sample t test. Counting data were measured by chi-square test and expressed as \u003cem\u003en\u003c/em\u003e/%. For consistency test, binary classification parameters were tested by Kappa test, and ordinal multi-classification parameters were tested by Weighted-Kappa test.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eRisk factors for predicting Ki-67 expression in peripheral lung cancer\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eKi-67 expression level was divided into high expression group and low expression group in training cohort and verification group. Logistic regression analysis of patient in gender, age, smoking and radiomics scores showed that there were no significant differences in age, gender and smoking between the training cohort and the verification group (\u003cem\u003eP\u003c/em\u003e\u0026gt;0.05). Radiomics scores were considered to be independent predictors of Ki-67 expression levels in peripheral lung cancer (Table 1).\u003c/p\u003e\n\u003cdiv align=\"center\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" align=\"left\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"5\" style=\"width: 30.9953%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTable 1\u003c/strong\u003e\u0026nbsp; \u0026nbsp;Prediction of risk factors for Ki-67 in peripheral lung cancer\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"23.41549295774648%\" rowspan=\"2\" style=\"width: 12.7887%;\"\u003e\n \u003cp\u003eVariables and intercepts\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"38.028169014084504%\" colspan=\"2\" valign=\"top\" style=\"width: 0.029%;\"\u003e\n \u003cp\u003eOne-factor Logistic regression\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"35.91549295774648%\" colspan=\"2\" valign=\"top\" style=\"width: 18.2581%;\"\u003e\n \u003cp\u003eMulti-factor Logistic regression\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.41013824884793%\" valign=\"top\" style=\"width: 17.9363%;\"\u003e\n \u003cp\u003eOR（95% CI）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"16.359447004608295%\" valign=\"top\" style=\"width: 7.8714%;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u0026nbsp;\u003c/em\u003evalue\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"32.25806451612903%\" valign=\"top\" style=\"width: 12.6278%;\"\u003e\n \u003cp\u003eOR（95% CI）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"14.516129032258064%\" valign=\"top\" style=\"width: 5.6302%;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u0026nbsp;\u003c/em\u003evalue\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"23.45679012345679%\" valign=\"top\" style=\"width: 12.7887%;\"\u003e\n \u003cp\u003eGender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.573192239858905%\" valign=\"top\" style=\"width: 17.9363%;\"\u003e\n \u003cp\u003e0.693（0.135-3.553）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.522045855379188%\" valign=\"top\" style=\"width: 7.8714%;\"\u003e\n \u003cp\u003e0.660\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.691358024691358%\" valign=\"top\" style=\"width: 12.6278%;\"\u003e\n \u003cp\u003eNA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.11111111111111%\" valign=\"top\" style=\"width: 5.6302%;\"\u003e\n \u003cp\u003eNA\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"23.45679012345679%\" valign=\"top\" style=\"width: 12.7887%;\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.573192239858905%\" valign=\"top\" style=\"width: 17.9363%;\"\u003e\n \u003cp\u003e0.982（0.899-1.073）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.522045855379188%\" valign=\"top\" style=\"width: 7.8714%;\"\u003e\n \u003cp\u003e0.687\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.691358024691358%\" valign=\"top\" style=\"width: 12.6278%;\"\u003e\n \u003cp\u003eNA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.11111111111111%\" valign=\"top\" style=\"width: 5.6302%;\"\u003e\n \u003cp\u003eNA\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"23.45679012345679%\" valign=\"top\" style=\"width: 12.7887%;\"\u003e\n \u003cp\u003esmoking and\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.573192239858905%\" valign=\"top\" style=\"width: 17.9363%;\"\u003e\n \u003cp\u003e0.798（0.534-5.361）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.522045855379188%\" valign=\"top\" style=\"width: 7.8714%;\"\u003e\n \u003cp\u003e0.583\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.691358024691358%\" valign=\"top\" style=\"width: 12.6278%;\"\u003e\n \u003cp\u003eNA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.11111111111111%\" valign=\"top\" style=\"width: 5.6302%;\"\u003e\n \u003cp\u003eNA\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"23.41549295774648%\" valign=\"top\" style=\"width: 12.7887%;\"\u003e\n \u003cp\u003eRadiomics scores\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.528169014084508%\" valign=\"top\" style=\"width: 17.9363%;\"\u003e\n \u003cp\u003e3.935（4.469-53.495）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.5%\" valign=\"top\" style=\"width: 6.8367%;\"\u003e\n \u003cp\u003e\u0026lt;0.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"24.64788732394366%\" valign=\"top\" style=\"width: 12.6278%;\"\u003e\n \u003cp\u003e3.935（4.469-53.495）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"11.091549295774648%\" valign=\"top\" style=\"width: 5.6302%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eNote：OR：odds ratio；NA：not available.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRadiomics scores\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe Z-scores of 396 radiomics features were normalized, the dimensionality of the data was reduced by LASSO regression model, and the optimal \u0026lambda; value was obtained by cross-validation (Fig. 2). A total of 7 non-zero coefficient features were screened (Fig. 3). The CT arterial phase images and radiomics features of Ki-67 high and low expression groups were shown.(Figure. 4 and Figure. 5). Then the radiomics risk score was calculated using the radiomics signature. The radiomics scores of the Ki-67 high and low expression groups were -1.89\u0026plusmn;1.13 and 0.51\u0026plusmn;1.18 in the training cohort, with statistical differences (\u003cem\u003et\u003c/em\u003e=-7.89, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01), and -1.28\u0026plusmn;1.21 and 0.74\u0026plusmn;0.87 in the validation cohort, with statistical differences (\u003cem\u003et\u003c/em\u003e=-3.76, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRadiomics prediction model formula\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eKi-67 expression level was taken as the dependent variable, and the radiomics score of peripheral lung cancer was taken as the independent variable to be included in the binary Logistics regression. An independent predictor of Ki-67 expression in peripheral lung cancer was the imaging score. The prediction model was Y=1/ [1+exp (-Z)], Z=2.275 * Radiomics_score + 0.712.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRadiomics signature predict performance\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eROC analysis of radiomics signature showed that AUC in the training cohort was 0.844 (95%CI: 0.725-0.964), sensitivity was 93%, and specificity was 71%. The Hosmer-Lemeshow test results showed that the calibration was 0.709 (Fig. 6). In the validation cohort, AUC was 0.881 (95%CI: 0.756-0.954), sensitivity was 91%, specificity was 75%, and calibration was 0.950 (Fig. 7). In both cohorts, the radiomics signature could predict the level of Ki-67 expression well.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDecision curve analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe analysis of DCA images showed that when the threshold probability was between 3% and 63%, radiomics signature based on CT enhanced arterial phase images predicted Ki-67 expression in peripheral lung cancer better than all patients considered Ki-67 high expression and all patients considered Ki-67 low expression (Fig. 8).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eWith the increasing incidence of peripheral lung cancer, early screening, prediction and prognosis of this disease are particularly important[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Therefore, when peripheral lung cancer is found, it has certain value for subsequent treatment to judge its malignancy degree and prognosis by using certain radiography examination and biological markers. Although the CT findings of peripheral lung cancer can detect nodules or masses in the chest, due to the relatively complex image features, it is difficult to identify benign and malignant, so it is necessary to diagnose peripheral lung cancer by radiography, tumor biological markers, pathological tissue section staining and other methods[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Ki-67, as a biomarker related to cell proliferation[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], is one of the most studied markers of proliferating cells. Studies[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] had shown that Ki-67 proliferation index was significantly correlated with the degree of differentiation and invasiveness of tumors. The higher the expression level of Ki-67, the faster the tumor growth, the higher the malignancy, the stronger the invasiveness, and the worse the prognosis, so it could be used to assess the degree of malignancy of the tumor. Ki-67 also plays an important role in the selection of tumor treatment and prognosis evaluation. Because of its convenient and non-invasive advantages, enhanced CT has become a common radiography examination method in clinical diagnosis of lung cancer. Different lung cancers have different growth patterns, so their CT findings show different imaging features. The expression of Ki-67 and other tumor markers determines the tumor growth pattern. Since the concept of \"radiomics\" was first proposed by Gillies et al.[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] in 2010 and subsequently improved by Lambin et al.[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], it had gradually developed into a research hotspot in the field of radiography. By analyzing images using radiomics method, molecular level information contained in images could be obtained. It could assist clinicians to make independent treatment plans for patients[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Chest enhanced CT images were analyzed by radiomics, and different image features had significant effect in predicting the expression level of molecular biological marker Ki-67, and could guide clinical treatment and prognosis.\u003c/p\u003e \u003cp\u003eIn this study, radiomics signature based on enhanced CT arterial phase images were used to predict the expression level of Ki-67 in peripheral lung cancer. The results of this study were as follows: The ROC curve of radiomics signature in predicting Ki-67 expression level of peripheral lung cancer showed that the AUC in the training cohort was 0.844, and the AUC in the validation cohort was 0.881. In the two test cohorts, radiomics signature could well predict the level of Ki-67 expression in peripheral lung cancer, and then predict the malignancy and prognosis of the tumor. It provided some guiding value for clinicians to formulate the next treatment measures. In this study, A.K software was used to extract lung cancer radiomics features, and then LASSO regression was used to further reduce the dimension of the feature datas and establish the model. Certain penalty terms were added to the cost function to make the established model have a sharper function, so it was not easy to produce certain overfitting. The optimal λ value was obtained by cross-validation, and the radiomics prediction model was closer to the ideal model. Previous studies had shown that radiomics score was an independent risk factor for predicting malignant tumors[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. In this study, only radiomics score was an independent risk factor for predicting Ki-67 expression based on the radiomics prediction model, which was consistent with previous studies. Some studies had shown that the expression level of Ki-67 was closely related to clinical data (gender, age, and smoking)[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e], but in this study, there was no statistical significance in age, gender, and smoking between the groups with high or low expression of Ki-67 (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05), which contradicted the results of previous studies and may be related to the small number of cases in this study. Some studies had found that Ki-67 was an independent risk factor for recurrence in patients with early lung adenocarcinoma after segmentation resection[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Therefore, it is very important for the selection of surgical methods to determine the expression of Ki-67 by non-invasive detection before surgery. In this study, we found that the radiomics signature can predict the expression level of Ki-67 noninvasively, which can guide clinicians to choose the appropriate surgical method.\u003c/p\u003e \u003cp\u003eThis study has the following limitations: First, the number of cases is small, so there is bias in this study. Second, this study takes peripheral lung cancer as the research object and has not included specific pathological types. Third, this study only used enhanced CT arterial phase images for feature extraction, and subsequently include other phases in the study. Fourth, some cases in this study were confirmed by percutaneous puncture biopsy, and although it can also be used to predict Ki-67 expression level, small biopsy tissues cannot represent the tumor tissue itself, so it is still challenging to predict Ki-67 expression level. Therefore, in the follow-up study, increasing the sample size, adding specific pathological types and detailed clinical and image-related information will help improve the detection performance of radiomics signature.\u003c/p\u003e \u003cp\u003eIn summary, this study showed that it has a high diagnostic value in predicting the Ki-67 expression level of peripheral lung cancer based on the radiomics signature created by enhanced CT arterial phase images, and provides a non-invasive examination means to assist clinicians in making a dedicated treatment plan for patients and evaluating prognosis.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.806509945750452%\" valign=\"top\"\u003e\n \u003cp\u003eAUC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"81.19349005424955%\"\u003e\n \u003cp\u003earea under the curve\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.806509945750452%\" valign=\"top\"\u003e\n \u003cp\u003eROC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"81.19349005424955%\"\u003e\n \u003cp\u003ereceiver operator characteristic\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.806509945750452%\" valign=\"top\"\u003e\n \u003cp\u003eDCA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"81.19349005424955%\"\u003e\n \u003cp\u003edecision curve analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.806509945750452%\" valign=\"top\"\u003e\n \u003cp\u003eVOI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"81.19349005424955%\"\u003e\n \u003cp\u003evolume of interest\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.806509945750452%\" valign=\"top\"\u003e\n \u003cp\u003eGLCM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"81.19349005424955%\"\u003e\n \u003cp\u003egray-level co-occurrence matrix\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.806509945750452%\" valign=\"top\"\u003e\n \u003cp\u003eGLRLM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"81.19349005424955%\"\u003e\n \u003cp\u003egray-level run length matrix\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"18.806509945750452%\" valign=\"top\"\u003e\n \u003cp\u003eGLSZM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"81.19349005424955%\"\u003e\n \u003cp\u003egray-level size zone matrix\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe sincerely appreciate all the patients who participated in this study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the Youth Innovation Team Scientific Research project of Shaanxi Provincial Education Department (grant no. 23JP035).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData and materials generated or analyzed during the study are available from the corresponding author by request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the ethics committee of Affiliated Hospital of Shaanxi University of Chinese Medicine, and written informed consent was obtained from all participants.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to publish\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors do not have any conflicts of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors contributed to the study\u0026apos;s conception and design. Material preparation, data collection, and analysis were performed by Wei Wei, Jian Geng, Dong Han, Yongjun Jia, Nan Yu, Shan Dang, Yong Yu, Yunsong Zheng and Lihua Fan. The first draft of the manuscript was written by Wei Wei, Lihua Fan and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eThai AA, Solomon BJ, Sequist LV, Gainor JF, Heist RS. Lung cancer. Lancet. 2021;398:535\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eK W JR. Y X, M Z, T R, Z G, Simulation framework for generating intratumor heterogeneity patterns in a cancer cell population. Eur J Surg oncology: J Eur Soc Surg Oncol Br Association Surg Oncol. 2023;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWm I. H I. Simulation framework for generating intratumor heterogeneity patterns in a cancer cell population. PLoS ONE. 2017;12.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eT KA, Ma RAVIPPC. B, Circulating Tumour Cells: Detection and Application in Advanced Non-Small Cell Lung Cancer. Int J Mol Sci. 2023;24.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEec de SW, Je J, van A I T. I C, J P, Decision Support Systems in Oncology. JCO Clin cancer Inf. 2019;3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJkt D, Cf F. The Role of Histologic Grading and Ki-67 Index in Predicting Outcomes in Pulmonary Carcinoid Tumors. Am J Surg Pathol. 2020;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKobara H, Mori H, Rafiq K, Fujihara S, Nishiyama N, Chiyo T, et al. Analysis of the amount of tissue sample necessary for mitotic count and Ki-67 index in gastrointestinal stromal tumor sampling. Oncol Rep. 2015;33:215\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZp M, Xl L, Tl KG, Hd Z, Yx W. Z. Application of radiomics based on chest CT-enhanced dual-phase imaging in the immunotherapy of non-small cell lung cancer. J X-Ray Sci Technol. 2023;31.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eD U-G AA. N B, M D. Expression of Ki-67 and Estrogen Receptor Beta in Primary Cutaneous Melanoma as a Potential Indicator of Regional Lymph Node Positivity. Applied immunohistochemistry \u0026amp; molecular morphology: AIMM. 2019;27.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMa M, Mr AS, Eh H, Pk C Jr. M. Association of distant recurrence-free survival with algorithmically extracted MRI characteristics in breast cancer. J Magn Reson imaging: JMRI. 2019;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNm B, C MEPP. D, H G, P T, Intratumoral and peritumoral radiomics for the pretreatment prediction of pathological complete response to neoadjuvant chemotherapy based on breast DCE-MRI. Breast cancer research: BCR. 2017;19.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBr L, Mk S, Br H, M Z, Ac N, Sc C. Renal mass biopsy\u0026ndash;a renaissance? J Urol. 2008;179.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMc AM Jr. M, H R, G L, K S, Probability of cancer in pulmonary nodules detected on first screening CT. N Engl J Med. 2013;369.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eC P, Cm RG et al. G, A V, M P, F G,. CT-guided transthoracic needle biopsy: advantages in histopathological and molecular tests. Future oncology (London, England). 2020;16.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMz M, Ah N, N N. MCM \u0026ndash;\u0026thinsp;2 and Ki \u0026ndash;\u0026thinsp;67 as proliferation markers in renal cell carcinoma: A quantitative and semi - quantitative analysis. Int Braz J Urol. 2016;42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRg B, Gm C, Ac G, Ct C, An C, Am A et al. Histologic correlation of expression of Ki-67 in squamous cell carcinoma of the glottis according to the degree of cell differentiation. Braz J Otorhinolaryngol. 2014;80.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRj G, Pe K. H H. Radiomics: Images Are More than Pictures, They Are Data. Radiology. 2016;278.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eE R-V PL, Rg van RLSC et al. S, P G,. Radiomics: extracting more information from medical images using advanced feature analysis. European journal of cancer (Oxford, England: 1990). 2012;48.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQ MZ, L Y CYG. Y T, C L, Development and validation of CT-based radiomics nomogram for the classification of benign parotid gland tumors. Med Phys. 2023;50.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eH X, Y QL. Z, J H, Y S, M L, Noninvasive evaluation of neutrophil extracellular traps signature predicts clinical outcomes and immunotherapy response in hepatocellular carcinoma. Front Immunol. 2023;14.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQ F, Sl L, Dp H, Yb H, Xj L, Z Z et al. CT Radiomics Model for Predicting the Ki-67 Index of Lung Cancer: An Exploratory Study. Front Oncol. 2021;11.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eS Y TM, K T MM. M C, S Y, Ki-67 labeling index is associated with recurrence after segmentectomy under video-assisted thoracoscopic surgery in stage I non-small cell lung cancer. Annals Thorac Cardiovasc surgery: official J Association Thorac Cardiovasc Surg Asia. 2011;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, Ki-67 proliferation index, Peripheral lung cancer, Tomography, X-ray computed","lastPublishedDoi":"10.21203/rs.3.rs-4323989/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4323989/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eThe association between radiomics features of peripheral lung cancer and Ki-67 expression level remains unclear. To develop a radiomics signature based on enhanced CT arterial phase image to estimate the expression situation of Ki-67 in peripheral lung cancer.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eA total of 117 peripheral lung cancer patients, who underwent contrast-enhanced CT scan in our hospital from May 2016 to November 2019, including 43 males and 74 females, aged 35 to 79 years old (median 54 years old). All the peripheral lung cancers were confirmed by histopathological and took in Ki-67 expression situation detection within 2 weeks after CT inspection, including 63 cases of Ki-67 low expression and 54 cases of Ki-67 high expression, which were retrospectively analyzed and were divided into training (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;82) and validation cohorts (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;35) in a ratio of 7:3. ITK-SNAP was used to manually outline the total tumor volume data of lung cancer on CT arterial phase images, and the radiomics features were extracted by A.K software. LASSO regression model was used to further screen features and construct radiomics labels, and the radiomics score of each patient was calculated, and then multi-factor logistic regression analysis was performed combined with clinical information to screen out independent risk factors for predicting Ki-67 levels. The predictive accuracy of the radiomics signature was quantified by the area under the curve (AUC) of receiver operator characteristic (ROC) curve in both the training and validation cohorts. The Hosmer-Lemeshow test was performed to evaluate the calibration degree of the radiomics. We performed decision curve analysis (DCA) to assess the clinical usefulness of the radiomics signature.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eSeven radiomics features were chosen from 396 candidate features to build a radiomics label that significantly correlated with Ki-67 expression level. The model showed good calibration and discrimination in the training cohort, with an AUC of 0.844 (95%CI: 0.725\u0026ndash;0.964), sensitivity of 93% and specificity of 71%, calibration degree of 0.709. In the validation cohort, AUC was 0.881 (95%CI: 0.756\u0026ndash;0.954), sensitivity was 91%, and specificity was 75%, calibration degree of 0.950. Univariate logistic regression analysis showed that there were no conspicuous differences in gender, age and smoke between the high and low Ki-67 expression (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). Using multivariate logistic regression model, radiomics signature were considered to be independent predictor of Ki-67 expression level in peripheral lung cancer. DCA for the radiomics signature in the training cohort showed that if the threshold probability was between 0.03 and 0.63, then using the radiomics signature to predict Ki-67 expression situation added more benefit than treating either all or no patients.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThe radiomics signature based on enhanced CT arterial phase image is benefitial to predict the expression of Ki-67 in peripheral lung cancer, which can assess the invasiveness and prognosis for peripheral lung cancer noninvasively.\u003c/p\u003e","manuscriptTitle":"CT study of radiomics features predicting Ki-67 expression level in peripheral lung cancer","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-09 02:20:59","doi":"10.21203/rs.3.rs-4323989/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":"4546ac76-cd0a-4bb6-ad1c-9def0c8bae12","owner":[],"postedDate":"May 9th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-10-14T07:09:14+00:00","versionOfRecord":[],"versionCreatedAt":"2024-05-09 02:20:59","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4323989","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4323989","identity":"rs-4323989","version":["v1"]},"buildId":"_2-kVJe1T_tPrBINL-cwx","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}