Prediction of Preterm Birth Based on Cervical Ultrasound Radiomics Combined with Clinical Features

Prediction of Preterm Birth Based on Cervical Ultrasound Radiomics Combined with Clinical Features | 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 Prediction of Preterm Birth Based on Cervical Ultrasound Radiomics Combined with Clinical Features HU Wenshu, ZHOU Chang, SUN Heng, LI Xinyi, XU Liang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7568405/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Objective: To evaluate the feasibility of predicting preterm birth using an ultrasound radiomics model combined with clinical features. Methods: We retrospectively analyzed 521 pregnant women who underwent prenatal care at Yichang Central People’s Hospital between January 2018 and August 2024. Patients were randomly assigned to a training set (n = 417) and a validation set (n = 104) at an 8:2 ratio. Radiomic features were extracted from the region of interest (ROI) of the cervix on 2D ultrasound images, and combined with clinical high-risk factors. All features were standardized and normalized to a (0,1) range. Feature selection was performed using variance thresholding, optimal feature selection (by number and percentage), and significance-based filtering. Logistic regression, random forest, and support vector machine models were constructed for preterm birth prediction. Model performance and clinical utility were evaluated using ROC curves, AUC, Hosmer-Lemeshow test, and decision curve analysis (DCA). Results: The combined model achieved the highest predictive performance for preterm birth, with AUCs of 0.874 and 0.841 in the training and validation sets, respectively, indicating good consistency. Hosmer-Lemeshow test and decision curves demonstrated good model calibration and high clinical net benefit. Conclusion: Ultrasound radiomics combined with clinical features can effectively predict preterm birth and may support early, non-invasive clinical interventions. Ultrasound Radiomics Preterm birth Cervix Machine learning Figures Figure 1 Figure 2 Figure 3 Figure 4 Background Preterm birth is a major global health concern, with an estimated 15 million infants born prematurely each year, accounting for approximately 11% of all live births [ 1 ]. It is a leading cause of neonatal mortality and long-term morbidity, potentially resulting in neurodevelopmental impairment, respiratory complications, and motor dysfunction. Spontaneous preterm birth accounts for nearly 50% of cases, and mid-trimester transvaginal cervical length < 25 mm is widely recognized as a key predictor for identifying high-risk pregnancies [ 2 , 3 ]. However, the predictive accuracy of single clinical or imaging parameters remains limited due to the multifactorial nature of preterm birth. Emerging evidence suggests that integrating multidimensional clinical and imaging features can improve risk stratification. Radiomics, a high-throughput technique for extracting quantitative imaging features, enables the identification of subtle tissue heterogeneity that is imperceptible to the human eye. When combined with relevant clinical data, radiomics may enhance the predictive performance of early risk assessment. While prior studies have primarily focused on the application of radiomics in cervical cancer diagnosis and prognosis [ 4 , 5 ]., its role in predicting obstetric complications, including preterm birth, remains underexplored. Recent advances in machine learning offer powerful tools to integrate complex imaging and clinical datasets for predictive modeling. Machine learning-based models have been shown to outperform conventional statistical methods in risk stratification, yet robust and clinically applicable models for preterm birth prediction are still limited. Therefore, this study aimed to extract cervical ultrasound radiomics features, integrate them with clinical and sonographic parameters, and develop a machine learning-based predictive model for preterm birth. This approach may facilitate early, non-invasive risk assessment and provide a practical reference for individualized clinical interventions. Methods Patient selection and data acquisition A retrospective analysis was conducted on the clinical and ultrasound data of pregnant women who attended the Maternal-Fetal Medicine Department of [Hospital Name] between January 2021 and July 2023. A total of 521 participants were enrolled based on predefined inclusion and exclusion criteria, including 238 women in the preterm group and 283 in the term group. This study was approved by the Institutional Review Board of Yichang Central People's Hospital, China Three Gorges University, with a waiver of informed consent (ethical approval number: 2024-11-14). The inclusion criteria were as follows: ① Singleton pregnancy; ② Gestational age between 16 and 28 weeks at the time of ultrasound examination; ③Availability of complete clinical records and ultrasound images of the cervix. The exclusion criteria were as follows: ① Multiple gestations; ② Congenital uterine malformations; ③ Severe maternal systemic diseases (e.g., cardiovascular, renal, or autoimmune disorders); ④ Incomplete clinical or imaging data; ⑤History of cervical surgery or trauma. Ultrasound Image Acquisition and Preprocessing Transvaginal ultrasound examinations were performed on pregnant women between 16 and 28 weeks of gestation using GE Voluson E8, GE Voluson E10, or Mindray Resona R9 systems equipped with RIC5-9-D or V11-3Hu volumetric probes (5–9 MHz). Participants were instructed to empty their bladder and positioned in the lithotomy position. The probe, covered with a condom, was gently inserted into the vagina, and imaging planes were adjusted to clearly visualize the internal os, external os, and endometrial line. Cervical length was measured, and the presence of internal os dilation was assessed. Two experienced obstetric sonographers (≥8 years of experience) independently delineated regions of interest (ROIs) on the anterior (ROI1) and posterior (ROI2) lips of the internal os. All ROIs were reviewed by a senior sonographer (10 years of experience), and discrepancies were resolved through consensus. The sonographers were blinded to the pregnancy outcomes. Images were stored in DICOM format for subsequent radiomic analysis. Clinical Parameter Collection Clinical data were retrospectively collected from medical records, including maternal age, body mass index (BMI), gravidity, parity, history of preterm birth, history of miscarriage, and use of assisted reproductive technology. These variables were coded for analysis, and any missing or ambiguous information was cross-verified by the research team. The collected clinical parameters were combined with extracted radiomic features to construct predictive models for preterm birth, ensuring both reproducibility and clinical relevance. Feature extraction, feature selection, and modeling All ultrasound images were independently delineated by two experienced obstetric ultrasonographers, and regions of interest (ROIs) were manually defined on the anterior lip (ROI1) and posterior lip (ROI2) of the cervical internal os (Fig. 1). The delineated images were imported into the Beijing Medical Intelligent Research Platform, where 1,125 radiomic features were automatically extracted. Feature pre-processing was conducted using min–max normalization to scale all variables to the range of 0–1. Optimal feature selection was performed through a combination of variance thresholding, significance-based filtering, and percentage-based selection. Logistic regression, random forest, and support vector machine (SVM) algorithms were then applied to construct predictive models for preterm birth (PTB). The ROI1-only, ROI2-only, and combined ROI1+ROI2 models were established, with logistic regression demonstrating the best predictive performance. All feature selection and modeling procedures were independently verified by two data analysts to ensure reproducibility. Statistical analysis and model evaluation Statistical analyses were performed using SPSS 27.0 and R 4.3.1 software. Continuous variables with normal distribution were expressed as mean ± standard deviation (± SD) and compared using the independent-samples t-test, while categorical variables were expressed as frequencies (percentages) and compared using the chi-square ( χ ²) test. Multivariate logistic regression was applied to identify independent clinical and ultrasound predictors of PTB. The performance of each predictive model was assessed using receiver operating characteristic (ROC) curves and area under the curve (AUC) values in both training and validation datasets. Model calibration was evaluated using the Hosmer–Lemeshow goodness-of-fit test, and a P value > 0.05 indicated good calibration. Clinical utility was further assessed using decision curve analysis (DCA). A P value < 0.05 was considered statistically significant. Results 2.1 Clinical Characteristics and Ultrasound Parameters Among 521 participants, 238 were preterm and 283 were term. Preterm group had higher proportions of multiparity, cervical length ≤ 25 mm, and internal os dilation (P < 0.05) (Table 1 ). Table 1 Comparison of Clinical Characteristics Between Preterm and Term Groups Parameter Preterm (n = 238) Term (n = 283) t/χ² P value Age (years, mean ± SD) 32.63 ± 3.38 32.03 ± 3.60 1.956 0.051 BMI (kg/m², mean ± SD) 25.02 ± 2.82 24.96 ± 2.97 0.221 0.825 Parity 4.429 0.035 Primipara (n, %) 66 (27.7) 103 (36.4) Multipara (n, %) 172 (72.3) 180 (63.6) History of Preterm Birth 21.102 < 0.001 Yes (n, %) 40 (16.8) 13 (4.5) No (n, %) 198 (83.2) 270 (95.4) History of Miscarriage 0.998 0.318 Yes (n, %) 32 (13.4) 30 (10.6) No (n, %) 206 (86.6) 253 (89.4) Assisted Reproduction 0.002 0.962 Yes (n, %) 23 (9.7) 27 (9.5) No (n, %) 215 (90.3) 256 (90.5) Cervical Length ≤ 25 mm 47.820 < 0.001 Yes (n, %) 46 (19.3) 4 (1.4) No (n, %) 192 (80.7) 279 (98.6) Cervical Internal Os Dilation 7.331 0.007 Yes (n, %) 18 (7.6) 7 (2.5) No (n, %) 220 (92.4) 276 (97.5) 2.2 Multivariate Logistic Regression Analysis Preterm birth was the dependent variable; parity, history of preterm birth, miscarriage, assisted reproduction, cervical length, and internal os dilation were independent variables. Logistic regression identified history of preterm birth (OR = 2.690, 95% CI: 1.284–5.636, P = 0.009) and cervical length ≤ 25 mm (OR = 13.313, 95% CI: 4.628–38.300, P < 0.001) as independent risk factors (Table 2 ). Table 2 Multivariate logistic regression analysis of independent risk factors for preterm birth. Parameter β SE Wald χ ² P value OR 95% CI Parity 0.021 0.211 0.010 0.921 1.021 0.675–1.544 History of Preterm Birth 0.990 0.377 6.879 0.009 2.690 1.284–5.636 History of Miscarriage 0.019 0.311 0.004 0.951 1.019 0.554–1.874 Assisted Reproduction 0.034 0.318 0.011 0.915 1.035 0.554–1.931 Cervical Length ≤ 25 mm 2.589 0.539 23.055 < 0.001 13.313 4.628–38.300 Cervical Internal Os Dilation 0.490 0.514 0.906 0.341 1.632 0.595–4.473 2.3 Radiomic Feature-Based Model and Evaluation Nine radiomic features combined with two clinical features (preterm birth history and cervical length ≤ 25 mm) were normalized and used to construct ROI1, ROI2, and ROI1 + ROI2 models. Logistic regression yielded the best performance. Training set AUCs: ROI1 0.826, ROI2 0.841, ROI1 + ROI2 0.874; validation set AUCs: ROI1 0.823, ROI2 0.772, ROI1 + ROI2 0.841 (Fig. 2 ). ROI1 + ROI2 model showed the highest diagnostic efficiency and good consistency (CI: training 0.874, validation 0.841). Hosmer-Lemeshow test confirmed good model calibration (Fig. 3 ). DCA showed clinical net benefit in the training set (Fig. 4 ). The Rad-score for the ROI1 + ROI2 model was calculated as follows: Rad-Score = 2.341×CL ≤ 25 -1.835×√GLRLM_GrayLevelNonUniformity_ROI1 -1.373×log-sigma-3-0-mm-3D_GL_RLM_GrayLevelNonUniformity_ROI2 -1.136×√GLRLM_GrayLevelNonUniformity_ROI2 -1.032×wavelet-LL_GL_RLM_GrayLevelNonUniformity_ROI2 -0.989×wavelet-LL_GL_RLM_GrayLevelNonUniformity_ROI1 -0.936×original_NG_TDM_Busyness_ROI1 -0.761×original_GL_RLM_GrayLevelNonUniformity_ROI1 -0.653×original_GL_RLM_GrayLevelNonUniformity_ROI2 + 1.653 Discussion Preterm birth results from the combined effects of multiple factors. Current domestic and international guidelines [ 3 , 6 ] recommend that women with a history of late miscarriage or preterm birth undergo transvaginal cervical length measurement at mid-gestation, with a cervical length ≤ 25 mm serving as a predictive indicator. However, interventions such as prolonged bed rest or cervical cerclage have not been shown to significantly reduce the incidence of preterm birth [ 7 , 8 ]. In nulliparous women, relying solely on clinical symptoms for preterm birth risk assessment may be insufficient and could even increase pregnancy-related risks, highlighting the need for more comprehensive and precise predictive tools for early clinical intervention. During pregnancy, cervical microstructural characteristics are closely associated with physiological changes. As gestation progresses, the rearrangement of collagen fibers and alterations in their composition directly affect cervical mechanical strength and compliance [ 9 , 10 ]. Ultrasound elastography provides a novel approach by quantitatively assessing cervical softening, offering potential for preterm birth risk prediction. Previous studies have demonstrated the feasibility of elastography in this context. For example, Chen et al. [ 11 ] analyzed clinical data and cervical elasticity parameters in 200 women at 6–8 weeks of gestation and identified the anterior lip cervical modulus and strain ratio as independent risk factors for preterm birth, constructing an early-pregnancy predictive model with favorable diagnostic performance. Similarly, Miao et al. [ 12 ] used E-cervix elastography to evaluate cervical tissue in 120 women with threatened preterm labor at 20–32 weeks of gestation, finding that the proportion of hard cervical tissue was significantly lower in the preterm group than in the term group [(35.75 ± 8.94)% vs. (61.30 ± 10.69)%, F = 156.88], and lower hard tissue proportion was moderately correlated with cervical shortening. These findings support the hypothesis that microstructural changes in the cervix (e.g., collagen remodeling, water content, and proteoglycan concentration) may precede macroscopic structural alterations [ 13 ]. Although ultrasound elastography shows promise as a noninvasive diagnostic tool, heterogeneity in instruments, technical approaches, patient selection, and study design limits its clinical application. In the present study, mid-gestation 2D ultrasound images were acquired using multiple devices and probes, and radiomic analysis was employed to extract deep cervical microstructural features. These features may quantitatively reflect cervical remodeling during pregnancy and its progression, demonstrating close association with preterm birth risk and providing a more precise quantitative basis for preterm birth prediction. In 1996, Iams et al. [ 14 ] first demonstrated in a multicenter prospective study that a short cervix measured by transvaginal ultrasound serves as an independent predictor of preterm birth. Subsequent studies have reported that funneling of the internal cervical os often occurs prior to preterm birth [ 15 ]. The histological characteristics of the internal cervical os differ significantly from those of the external os: while the external os is primarily composed of collagen fibers, the internal os contains approximately 50–60% smooth muscle. These smooth muscle bundles encircle the internal os, maintaining cervical closure during pregnancy and exhibiting enhanced contractile capacity under oxytocin stimulation. Some researchers have proposed that the internal os possesses a specialized sphincter-like structure, which effectively ensures fetal retention in utero during gestation, and initiates labor only when its functional threshold is exceeded [ 16 ]. This concept provides a clearer understanding of the role of the cervix in maintaining pregnancy and explains how cervical insufficiency can lead to preterm birth. In a study investigating the microstructure of the cervical internal os in pregnant rats [ 17 ], significant rearrangement of collagen fibers was observed in the internal os region during gestation, whereas other cervical regions showed minimal changes. As pregnancy progressed, collagen fibers in the internal os gradually shifted from an orientation parallel to the cervical canal to a vertical arrangement, which may influence the degree of internal os dilation and its biomechanical properties. Furthermore, when collagen fibers were parallel to the cervical canal, the radial displacement of the anterior and posterior lips of the cervix increased approximately 6.5-fold compared with the vertical orientation, indicating a direct relationship between collagen fiber remodeling and funneling of the internal os. Therefore, focusing on the internal cervical os can enhance understanding of the mechanisms maintaining pregnancy and the pathophysiology of preterm birth, providing more precise predictive indicators and potential intervention strategies. In the present study, the combined radiomic features of the anterior (ROI1) and posterior (ROI2) lips of the internal os were used to construct the ROI1 + ROI2 combined model, which demonstrated significant diagnostic performance in predicting preterm birth. This finding further supports the critical role of microstructural changes in the internal cervical os in both pregnancy maintenance and preterm birth prediction. In this study, a predictive model for preterm birth was established by integrating cervical ultrasound radiomic features with clinical parameters. The results demonstrated that a history of preterm birth and cervical length ≤ 25 mm were independent risk factors, consistent with previous reports [ 18 , 19 ]. Cervical length, as a commonly used predictor, was further validated for its effectiveness in assessing preterm birth risk. Nine radiomic features were extracted from ultrasound images, including various high-order features reflecting tissue texture and gray-level distribution. Among them, wavelet-LL_glrlm_GrayLevelNonUniformity and log-sigma-3-0-mm-3D_glrlm_GrayLevelNonUniformity showed the strongest associations with preterm birth; these features were derived from the internal and external cervical os regions, reflecting gray-level heterogeneity and subtle textural differences, respectively. Except for the original shape feature (original_shape_MinorAxisLength), all other features were high-order, capturing microstructural differences that are difficult to detect with conventional visual assessment, indicating the potential value of radiomic features for noninvasive, early prediction. Furthermore, decision curve analysis revealed that the combined model could substantially improve clinical benefit in practice, supporting its potential applicability. Our findings suggest that cervical ultrasound radiomic features can significantly enhance the prediction of preterm birth. By integrating multidimensional radiomic features with clinical data, the constructed model not only improves risk identification but also addresses the limitations of traditional single-parameter assessment. Limitations This study is limited by its retrospective design and relatively small sample size, which may affect generalizability. Future prospective, multicenter, and larger-scale studies are warranted. Other factors such as maternal health and environmental influences should also be explored. This model requires validation in diverse populations and multicenter cohorts to confirm broader applicability. Conclusion Cervical ultrasound radiomics combined with clinical features effectively predicts preterm birth. This model improves early risk identification and supports timely clinical intervention, offering a promising tool for personalized management strategies. Abbreviations PTB — Preterm Birth CL — Cervical Length ROI — Region of Interest ROI1 — Cervical Internal Os Anterior Lip Region of Interest ROI2 — Cervical Internal Os Posterior Lip Region of Interest Rad-score — Radiomics Score GLRLM — Gray Level Run Length Matrix ROC — Receiver Operating Characteristic AUC — Area Under the Curve DCA — Decision Curve Analysis SVM — Support Vector Machine LR — Logistic Regression BMI — Body Mass Index SD — Standard Deviation DICOM — Digital Imaging and Communications in Medicine ICC — Intraclass Correlation Coefficient ESC — European Society of Cardiology SPSS — Statistical Package for the Social Sciences R — R Statistical Software OR — Odds Ratio CI — Confidence Interval US — Ultrasound Declarations Ethics approval and consent to participate This study was approved by the Ethics Committee of Yichang Central People's Hospital, China Three Gorges University (approval number: 2024-11-14). Written informed consent was obtained from all participants prior to enrollment. Consent for publication Not applicable. Availability of data and materials The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request. Competing interests The authors declare that they have no competing interests. Funding No funding was received for this study. Authors' contributions W.H. and C.Z. conceived and designed the study. W.H. collected and organized the clinical and imaging data. H.S. and L.X. performed the radiomics feature extraction and statistical analyses. X.L. contributed to the interpretation of results and drafting of the manuscript. W.H. wrote the initial draft, and C.Z. critically revised the manuscript for important intellectual content. All authors reviewed and approved the final version of the manuscript. Acknowledgements The authors would like to thank the staff of the Department of Ultrasound, The First College of Clinical Medical Science, China Three Gorges University & Yichang Central People's Hospital, for their support in data collection and technical assistance. References Vogel J P, Chawanpaiboon S, Moller A, et al. The global epidemiology of preterm birth[J]. Best Pract Res Clin Obstet Gynaecol. 2018,52:3-12. WU P L，XUE Q，LIU X X，SUN X. Research progress on cervical cerclage for preventing preterm birth in twin pregnancies. Chinese Journal of Perinatal Medicine, 2024,27(3):258-261.(in Chinese). ZHONG H Y X H. Clinical guidelines for the prevention and treatment of preterm birth (version 2024). Chinese Journal of Obstetrics and Gynecology, 2024,59(04):257-269.(in Chinese). Xin W, Rixin S, Linrui L, Zhihui Q, Long L, Yu Z. Machine learning-based radiomics for predicting outcomes in cervical cancer patients undergoing concurrent chemoradiotherapy. Comput Biol Med. 2024;177:108593. Qiu H, Wang M, Wang S, et al. Integrating MRI-based radiomics and clinicopathological features for preoperative prognostication of early-stage cervical adenocarcinoma patients: in comparison to deep learning approach. Cancer Imaging. 2024;24(1):101. The Lancet. The unfinished agenda of preterm births[J]. Lancet. 2016;388(10058):2323. Saccone G, Maruotti G M, Morlando M, et al. Randomized trial of screening for preterm birth in low-risk women - the preterm birth screening study[J]. Int J Nurs Stud , 2024,6(5):101267. XU G C，ZHANG J P. Controversy and views on cervical cerclage. Chinese Journal of Perinatal Medicine, 2023,26(10):793-796.(in Chinese). Moghaddam A O, Lin Z, Sivaguru M, et al. Heterogeneous microstructural changes of the cervix influence cervical funneling[J]. Acta Biomaterialia, 2022,140:434-445. Stone J, House M. Measurement of cervical softness before cerclage placement with an aspiration-based device[J]. Am J Obstet Gynecol MFM , 2023,5(4):100881. SIHAN C，BING H U，XIAOZHEN X，LING D，YING Y. Construction of Prediction Model for Preterm Birth in Early Pregnancy via Cervical Elastography. Chinese Journal of Medical Imaging, 2023,31(12):1298-1303.(in Chinese). MOU Y T，LONG Y，LUO Y C，ZHANG S. Application of E-cervix elastography technology in pregnant women with threatened preterm birth: an analysis of 120 cases. Chinese Journal of Perinatal Medicine, 2024,27(08):656-661.(in Chinese). Patberg ET, Wells M, Vahanian SA, et al. Use of cervical elastography at 18 to 22 weeks' gestation in the prediction of spontaneous preterm birth[J]. Am J Obstet Gynecol. 2021;225(5):525.e1-525.e9. Iams J D, Goldenberg R L, Meis P J, et al. The length of the cervix and the risk of spontaneous premature delivery. National Institute of Child Health and Human Development Maternal Fetal Medicine Unit Network[J]. N Engl J Med, 1996,334(9):567-572. Coutinho C M, Sotiriadis A, Odibo A, et al. ISUOG Practice Guidelines: role of ultrasound in the prediction of spontaneous preterm birth[J]. Ultrasound Obstet Gynecol. 2022,60(3):435-456. Marinescu P S, Young R C, Miller L A, et al. Mid-trimester uterine electromyography in patients with a short cervix[J]. Am J Obstet Gynecol, 2022,227(1):81-83. Moghaddam A O, Lin Z, Sivaguru M, et al. Heterogeneous microstructural changes of the cervix influence cervical funneling[J]. Acta Biomaterialia, 2022,140:434-445. Hessami K, D'Alberti E, Mascio D D, et al. Universal cervical length screening and risk of spontaneous preterm birth: a systematic review and meta-analysis[J]. Am J Obstet Gynecol MFM, 2024,6(5S):101343. Hessami K, D'Alberti E, Mascio D D, et al. Universal cervical length screening and risk of spontaneous preterm birth: a systematic review and meta-analysis[J]. Am J Obstet Gynecol MFM, 2024,6(5S):101343. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 24 Oct, 2025 Reviewers agreed at journal 14 Oct, 2025 Reviewers invited by journal 06 Oct, 2025 Editor invited by journal 11 Sep, 2025 Editor assigned by journal 10 Sep, 2025 Submission checks completed at journal 10 Sep, 2025 First submitted to journal 08 Sep, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7568405","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":530858600,"identity":"7293734e-1abe-45ca-9cb8-3e2e7710b88c","order_by":0,"name":"HU Wenshu","email":"","orcid":"","institution":"China Three Gorges University \u0026 Yichang Central People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"HU","middleName":"","lastName":"Wenshu","suffix":""},{"id":530858601,"identity":"e9d3d4d1-cf08-4d08-ac42-06ea7ce30f77","order_by":1,"name":"ZHOU Chang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA00lEQVRIiWNgGAWjYDACCSBmbLDhYWM4fODAhwritaTJ8TMeSzw44wzxWg4bSzafMT7M20KEDvnZzcckfu5gTtxw7MyHA7wNDPL8Ygfwa2GccyxNsvcMW+KGM2c3HJDcwWA4c3YCfi3MEjlm0oxtPIkbbgC1GJ5hSDC4TUALG0SLROKG+28eHEhsI0ILD0SLgbFkwxmGAweJ0SIhkZZs2duWIMfPcMzgYMMZCcJ+kZ+RfPDGz7b/oKh8/PlPhY08vzQBLRi2kqZ8FIyCUTAKRgF2AAAIaEsweo4HDQAAAABJRU5ErkJggg==","orcid":"","institution":"China Three Gorges University \u0026 Yichang Central People's Hospital","correspondingAuthor":true,"prefix":"","firstName":"ZHOU","middleName":"","lastName":"Chang","suffix":""},{"id":530858602,"identity":"21ea83aa-9493-4527-bdec-408b367a6c74","order_by":2,"name":"SUN Heng","email":"","orcid":"","institution":"China Three Gorges University \u0026 Yichang Central People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"SUN","middleName":"","lastName":"Heng","suffix":""},{"id":530858603,"identity":"14765585-1c25-4944-a2ed-c2cb552d2195","order_by":3,"name":"LI Xinyi","email":"","orcid":"","institution":"China Three Gorges University \u0026 Yichang Central People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"LI","middleName":"","lastName":"Xinyi","suffix":""},{"id":530858605,"identity":"777bf82b-da5b-4968-b3a1-f4858b06c8f1","order_by":4,"name":"XU Liang","email":"","orcid":"","institution":"China Three Gorges University \u0026 Yichang Central People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"XU","middleName":"","lastName":"Liang","suffix":""}],"badges":[],"createdAt":"2025-09-09 02:23:06","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7568405/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7568405/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":93778938,"identity":"58d3f511-605c-41b9-8169-80dadcbacd24","added_by":"auto","created_at":"2025-10-17 12:50:27","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":753637,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/4cf9f82096db2fee6e22918c.docx"},{"id":93778935,"identity":"1ec10170-6315-47a9-b63b-4672b0eb538b","added_by":"auto","created_at":"2025-10-17 12:50:27","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6698,"visible":true,"origin":"","legend":"","description":"","filename":"99841446ecc5446ab6d2970e3e31c04a.json","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/9f925d2223e3bf71599636b9.json"},{"id":93777664,"identity":"ec26e0ac-ad2a-4a08-ab9a-59212b4a0f80","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":70754,"visible":true,"origin":"","legend":"","description":"","filename":"99841446ecc5446ab6d2970e3e31c04a1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/52d0034c7719ca34f527f687.xml"},{"id":93780072,"identity":"3a50ce24-67c4-407f-bcdb-780902f92305","added_by":"auto","created_at":"2025-10-17 12:58:27","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":91439,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/f27126b2689d448ea5ab89d7.png"},{"id":93778939,"identity":"d4c50035-23a0-42d2-a489-3688294194d8","added_by":"auto","created_at":"2025-10-17 12:50:27","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":92601,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/9d8a1eaa7a55d8e01400ead6.png"},{"id":93777679,"identity":"588850c8-86c3-477c-bfcc-24e0bd2fa699","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":438315,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/884be6101bdfcbb298519c19.png"},{"id":93778936,"identity":"3bbe7977-c52c-4fbc-a77e-ab977fd63d51","added_by":"auto","created_at":"2025-10-17 12:50:27","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35936,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/49e3450e239ed00851933717.png"},{"id":93778943,"identity":"7740ce98-d090-4748-a9f0-5f0a4e44ef43","added_by":"auto","created_at":"2025-10-17 12:50:27","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53018,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/a51e435829d07ad9343c7808.png"},{"id":93777670,"identity":"49cf540a-07b9-45da-9d1d-7154aa442145","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":20547,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/3f4b14cce2a1fd132254b66b.png"},{"id":93778941,"identity":"d72476db-d5de-4289-a243-6e037b8cb3d2","added_by":"auto","created_at":"2025-10-17 12:50:27","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":20666,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/6bf99dae0a1f67bb994ea9cc.png"},{"id":93777678,"identity":"b14f36ff-597a-4cee-88a6-7bcc58b698b3","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":105041,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/3680aef6144193919fa49526.png"},{"id":93777674,"identity":"5bb7e0b4-d106-42df-861b-bcce665f8748","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11791,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/68d80c848fd998ba50aad034.png"},{"id":93778942,"identity":"906d2f12-1394-48a4-8022-b6899a28ce2c","added_by":"auto","created_at":"2025-10-17 12:50:27","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15582,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/f4f5429901d44164909af5f1.png"},{"id":93777676,"identity":"9236d0c1-1ae7-4436-9999-9fc2196dbf6d","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"xml","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":69495,"visible":true,"origin":"","legend":"","description":"","filename":"99841446ecc5446ab6d2970e3e31c04a1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/2a15f932c984546dc7f9c72a.xml"},{"id":93777680,"identity":"1ac361f6-8219-42bf-a61c-a904416583fb","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"html","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":76161,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/92d0c3b1a439056039f59971.html"},{"id":93777668,"identity":"e64c1bcb-300c-43ce-be40-d209201f6025","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":418419,"visible":true,"origin":"","legend":"\u003cp\u003eDelineation of regions of interest (ROIs) at the anterior and posterior lips of the cervical internal os.\u003c/p\u003e\n\u003cp\u003eNote: The red area represents ROI1, and the blue area represents ROI2.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/6d043adac3b164f4a214b903.png"},{"id":93780073,"identity":"2872ba04-5083-4916-a3f4-c02037bfad87","added_by":"auto","created_at":"2025-10-17 12:58:27","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":257038,"visible":true,"origin":"","legend":"\u003cp\u003eROC curves of ROI1, ROI2, and ROI1+ROI2 models in training and validation sets.\u003c/p\u003e\n\u003cp\u003eROC curve analysis of radiomics models for predicting the outcome. The graph compares the diagnostic performance of six models: RIO2 combined model in the validation set (orange curve, AUC: 0.772), RIO2 combined model in the training set (black curve, AUC: 0.841), RIO1 combined model in the validation set (magenta curve, AUC: 0.823), RIO1 combined model in the training set (brown curve, AUC: 0.826), RIO1+RIO2 combined model in the validation set (blue curve, AUC: 0.841), and RIO1+RIO2 combined model in the training set (red curve, AUC: 0.874).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/257275cb6857eb185db4038d.png"},{"id":93777662,"identity":"cd810c6e-0b38-4b3e-b4b8-1d2009493285","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":273027,"visible":true,"origin":"","legend":"\u003cp\u003eHosmer-Lemeshow calibration curves for ROI1+ROI2 model\u003c/p\u003e\n\u003cp\u003eNote: Blue represents the training set, and orange represents the validation set.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/1593d3d1f8bb9d36a183e439.png"},{"id":93777673,"identity":"88639635-43e4-4b6f-8fd4-af059885590d","added_by":"auto","created_at":"2025-10-17 12:42:27","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":222311,"visible":true,"origin":"","legend":"\u003cp\u003eDecision curve analysis for clinical net benefit of ROI1+ROI2 model.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/a7beb15e9b908780b24a26df.png"},{"id":93780085,"identity":"129fe039-3732-4322-a5dc-db4aa55d4be9","added_by":"auto","created_at":"2025-10-17 12:58:32","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1972745,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7568405/v1/4199b49d-33a9-4085-a2df-8651973468ba.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Prediction of Preterm Birth Based on Cervical Ultrasound Radiomics Combined with Clinical Features","fulltext":[{"header":"Background","content":"\u003cp\u003ePreterm birth is a major global health concern, with an estimated 15\u0026nbsp;million infants born prematurely each year, accounting for approximately 11% of all live births [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. It is a leading cause of neonatal mortality and long-term morbidity, potentially resulting in neurodevelopmental impairment, respiratory complications, and motor dysfunction. Spontaneous preterm birth accounts for nearly 50% of cases, and mid-trimester transvaginal cervical length \u0026lt; 25 mm is widely recognized as a key predictor for identifying high-risk pregnancies [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eHowever, the predictive accuracy of single clinical or imaging parameters remains limited due to the multifactorial nature of preterm birth. Emerging evidence suggests that integrating multidimensional clinical and imaging features can improve risk stratification. Radiomics, a high-throughput technique for extracting quantitative imaging features, enables the identification of subtle tissue heterogeneity that is imperceptible to the human eye. When combined with relevant clinical data, radiomics may enhance the predictive performance of early risk assessment. While prior studies have primarily focused on the application of radiomics in cervical cancer diagnosis and prognosis [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]., its role in predicting obstetric complications, including preterm birth, remains underexplored.\u003c/p\u003e\u003cp\u003eRecent advances in machine learning offer powerful tools to integrate complex imaging and clinical datasets for predictive modeling. Machine learning-based models have been shown to outperform conventional statistical methods in risk stratification, yet robust and clinically applicable models for preterm birth prediction are still limited. Therefore, this study aimed to extract cervical ultrasound radiomics features, integrate them with clinical and sonographic parameters, and develop a machine learning-based predictive model for preterm birth. This approach may facilitate early, non-invasive risk assessment and provide a practical reference for individualized clinical interventions.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003ePatient selection and data acquisition\u003c/p\u003e\n\u003cp\u003eA retrospective analysis was conducted on the clinical and ultrasound data of pregnant women who attended the Maternal-Fetal Medicine Department of [Hospital Name] between January 2021 and July 2023. A total of 521 participants were enrolled based on predefined inclusion and exclusion criteria, including 238 women in the preterm group and 283 in the term group. This study was approved by the Institutional Review Board of Yichang Central People\u0026apos;s Hospital, China Three Gorges University, with a waiver of informed consent (ethical approval number: 2024-11-14).\u003c/p\u003e\n\u003cp\u003eThe inclusion criteria were as follows:\u003c/p\u003e\n\u003cp\u003e①\u0026nbsp;\u0026nbsp;Singleton pregnancy;\u003c/p\u003e\n\u003cp\u003e② Gestational age between 16 and 28 weeks at the time of ultrasound examination;\u003c/p\u003e\n\u003cp\u003e③Availability of complete clinical records and ultrasound images of the cervix.\u003c/p\u003e\n\u003cp\u003eThe exclusion criteria were as follows:\u003c/p\u003e\n\u003cp\u003e①\u0026nbsp;\u0026nbsp;Multiple gestations;\u003c/p\u003e\n\u003cp\u003e② Congenital uterine malformations;\u003c/p\u003e\n\u003cp\u003e③ Severe maternal systemic diseases (e.g., cardiovascular, renal, or autoimmune disorders);\u003c/p\u003e\n\u003cp\u003e④ Incomplete clinical or imaging data;\u003c/p\u003e\n\u003cp\u003e⑤History of cervical surgery or trauma.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eUltrasound Image Acquisition and Preprocessing\u003c/p\u003e\n\u003cp\u003eTransvaginal ultrasound examinations were performed on pregnant women between 16 and 28 weeks of gestation using GE Voluson E8, GE Voluson E10, or Mindray Resona R9 systems equipped with RIC5-9-D or V11-3Hu volumetric probes (5\u0026ndash;9 MHz). Participants were instructed to empty their bladder and positioned in the lithotomy position. The probe, covered with a condom, was gently inserted into the vagina, and imaging planes were adjusted to clearly visualize the internal os, external os, and endometrial line. Cervical length was measured, and the presence of internal os dilation was assessed. Two experienced obstetric sonographers (\u0026ge;8 years of experience) independently delineated regions of interest (ROIs) on the anterior (ROI1) and posterior (ROI2) lips of the internal os. All ROIs were reviewed by a senior sonographer (10 years of experience), and discrepancies were resolved through consensus. The sonographers were blinded to the pregnancy outcomes. Images were stored in DICOM format for subsequent radiomic analysis.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;Clinical Parameter Collection\u003c/p\u003e\n\u003cp\u003eClinical data were retrospectively collected from medical records, including maternal age, body mass index (BMI), gravidity, parity, history of preterm birth, history of miscarriage, and use of assisted reproductive technology. These variables were coded for analysis, and any missing or ambiguous information was cross-verified by the research team. The collected clinical parameters were combined with extracted radiomic features to construct predictive models for preterm birth, ensuring both reproducibility and clinical relevance.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFeature extraction, feature selection, and modeling\u003c/p\u003e\n\u003cp\u003eAll ultrasound images were independently delineated by two experienced obstetric ultrasonographers, and regions of interest (ROIs) were manually defined on the anterior lip (ROI1) and posterior lip (ROI2) of the cervical internal os (Fig. 1). The delineated images were imported into the Beijing Medical Intelligent Research Platform, where 1,125 radiomic features were automatically extracted. Feature pre-processing was conducted using min\u0026ndash;max normalization to scale all variables to the range of 0\u0026ndash;1. Optimal feature selection was performed through a combination of variance thresholding, significance-based filtering, and percentage-based selection. Logistic regression, random forest, and support vector machine (SVM) algorithms were then applied to construct predictive models for preterm birth (PTB). The ROI1-only, ROI2-only, and combined ROI1+ROI2 models were established, with logistic regression demonstrating the best predictive performance. All feature selection and modeling procedures were independently verified by two data analysts to ensure reproducibility.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;Statistical analysis and model evaluation\u003c/p\u003e\n\u003cp\u003eStatistical analyses were performed using SPSS 27.0 and R 4.3.1 software. Continuous variables with normal distribution were expressed as mean\u0026nbsp;\u0026plusmn;\u0026nbsp;standard deviation (\u0026plusmn;\u0026nbsp;SD) and compared using the independent-samples t-test, while categorical variables were expressed as frequencies (percentages) and compared using the chi-square (\u003cem\u003e\u0026chi;\u003c/em\u003e\u0026sup2;) test. Multivariate logistic regression was applied to identify independent clinical and ultrasound predictors of PTB. The performance of each predictive model was assessed using receiver operating characteristic (ROC) curves and area under the curve (AUC) values in both training and validation datasets. Model calibration was evaluated using the Hosmer\u0026ndash;Lemeshow goodness-of-fit test, and a P value \u0026gt; 0.05 indicated good calibration. Clinical utility was further assessed using decision curve analysis (DCA). A P value \u0026lt; 0.05 was considered statistically significant.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Clinical Characteristics and Ultrasound Parameters\u003c/h2\u003e\u003cp\u003eAmong 521 participants, 238 were preterm and 283 were term. Preterm group had higher proportions of multiparity, cervical length\u0026thinsp;\u0026le;\u0026thinsp;25 mm, and internal os dilation (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eComparison of Clinical Characteristics Between Preterm and Term Groups\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eParameter\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePreterm (n\u0026thinsp;=\u0026thinsp;238)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTerm (n\u0026thinsp;=\u0026thinsp;283)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cem\u003et/χ\u0026sup2;\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge (years, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e32.63\u0026thinsp;\u0026plusmn;\u0026thinsp;3.38\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32.03\u0026thinsp;\u0026plusmn;\u0026thinsp;3.60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1.956\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.051\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBMI (kg/m\u0026sup2;, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e25.02\u0026thinsp;\u0026plusmn;\u0026thinsp;2.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e24.96\u0026thinsp;\u0026plusmn;\u0026thinsp;2.97\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.221\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.825\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eParity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.429\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.035\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrimipara (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e66 (27.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e103 (36.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMultipara (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e172 (72.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e180 (63.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHistory of Preterm Birth\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e21.102\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e40 (16.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e13 (4.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e198 (83.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e270 (95.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHistory of Miscarriage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.998\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.318\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e32 (13.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e30 (10.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e206 (86.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e253 (89.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAssisted Reproduction\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.002\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.962\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e23 (9.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e27 (9.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e215 (90.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e256 (90.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCervical Length\u0026thinsp;\u0026le;\u0026thinsp;25 mm\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e47.820\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e46 (19.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4 (1.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e192 (80.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e279 (98.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCervical Internal Os Dilation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e7.331\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.007\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e18 (7.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e7 (2.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo (n, %)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e220 (92.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e276 (97.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\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 Multivariate Logistic Regression Analysis\u003c/h2\u003e\u003cp\u003ePreterm birth was the dependent variable; parity, history of preterm birth, miscarriage, assisted reproduction, cervical length, and internal os dilation were independent variables. Logistic regression identified history of preterm birth (OR\u0026thinsp;=\u0026thinsp;2.690, 95% CI: 1.284\u0026ndash;5.636, P\u0026thinsp;=\u0026thinsp;0.009) and cervical length\u0026thinsp;\u0026le;\u0026thinsp;25 mm (OR\u0026thinsp;=\u0026thinsp;13.313, 95% CI: 4.628\u0026ndash;38.300, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001) as independent risk factors (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\u003eMultivariate logistic regression analysis of independent risk factors for preterm birth.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eParameter\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cem\u003eβ\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cem\u003eSE\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eWald \u003cem\u003eχ\u003c/em\u003e\u0026sup2;\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cem\u003eOR\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003e95% \u003cem\u003eCI\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eParity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.211\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.010\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.921\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.021\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.675\u0026ndash;1.544\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHistory of Preterm Birth\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.990\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.377\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e6.879\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.009\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e2.690\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e1.284\u0026ndash;5.636\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHistory of Miscarriage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.019\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.311\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.004\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.951\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.019\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.554\u0026ndash;1.874\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAssisted Reproduction\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.034\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.318\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.011\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.915\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.035\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.554\u0026ndash;1.931\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCervical Length\u0026thinsp;\u0026le;\u0026thinsp;25 mm\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2.589\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.539\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e23.055\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e13.313\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e4.628\u0026ndash;38.300\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCervical Internal Os Dilation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.490\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.514\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.906\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.341\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.632\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.595\u0026ndash;4.473\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=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Radiomic Feature-Based Model and Evaluation\u003c/h2\u003e\u003cp\u003eNine radiomic features combined with two clinical features (preterm birth history and cervical length\u0026thinsp;\u0026le;\u0026thinsp;25 mm) were normalized and used to construct ROI1, ROI2, and ROI1\u0026thinsp;+\u0026thinsp;ROI2 models. Logistic regression yielded the best performance. Training set AUCs: ROI1 0.826, ROI2 0.841, ROI1\u0026thinsp;+\u0026thinsp;ROI2 0.874; validation set AUCs: ROI1 0.823, ROI2 0.772, ROI1\u0026thinsp;+\u0026thinsp;ROI2 0.841 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). ROI1\u0026thinsp;+\u0026thinsp;ROI2 model showed the highest diagnostic efficiency and good consistency (CI: training 0.874, validation 0.841). Hosmer-Lemeshow test confirmed good model calibration (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). DCA showed clinical net benefit in the training set (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe Rad-score for the ROI1\u0026thinsp;+\u0026thinsp;ROI2 model was calculated as follows:\u003c/p\u003e\n\u003cp\u003eRad-Score\u0026thinsp;=\u0026thinsp;2.341\u0026times;CL\u0026thinsp;\u0026le;\u0026thinsp;25 -1.835\u0026times;\u0026radic;GLRLM_GrayLevelNonUniformity_ROI1\u003c/p\u003e\n\u003cp\u003e-1.373\u0026times;log-sigma-3-0-mm-3D_GL_RLM_GrayLevelNonUniformity_ROI2\u003c/p\u003e\n\u003cp\u003e-1.136\u0026times;\u0026radic;GLRLM_GrayLevelNonUniformity_ROI2\u003c/p\u003e\n\u003cp\u003e-1.032\u0026times;wavelet-LL_GL_RLM_GrayLevelNonUniformity_ROI2\u003c/p\u003e\n\u003cp\u003e-0.989\u0026times;wavelet-LL_GL_RLM_GrayLevelNonUniformity_ROI1\u003c/p\u003e\n\u003cp\u003e-0.936\u0026times;original_NG_TDM_Busyness_ROI1\u003c/p\u003e\n\u003cp\u003e-0.761\u0026times;original_GL_RLM_GrayLevelNonUniformity_ROI1\u003c/p\u003e\n\u003cp\u003e-0.653\u0026times;original_GL_RLM_GrayLevelNonUniformity_ROI2\u0026thinsp;+\u0026thinsp;1.653\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003ePreterm birth results from the combined effects of multiple factors. Current domestic and international guidelines [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e] recommend that women with a history of late miscarriage or preterm birth undergo transvaginal cervical length measurement at mid-gestation, with a cervical length\u0026thinsp;\u0026le;\u0026thinsp;25 mm serving as a predictive indicator. However, interventions such as prolonged bed rest or cervical cerclage have not been shown to significantly reduce the incidence of preterm birth [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. In nulliparous women, relying solely on clinical symptoms for preterm birth risk assessment may be insufficient and could even increase pregnancy-related risks, highlighting the need for more comprehensive and precise predictive tools for early clinical intervention.\u003c/p\u003e\u003cp\u003eDuring pregnancy, cervical microstructural characteristics are closely associated with physiological changes. As gestation progresses, the rearrangement of collagen fibers and alterations in their composition directly affect cervical mechanical strength and compliance [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Ultrasound elastography provides a novel approach by quantitatively assessing cervical softening, offering potential for preterm birth risk prediction. Previous studies have demonstrated the feasibility of elastography in this context. For example, Chen et al. [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] analyzed clinical data and cervical elasticity parameters in 200 women at 6\u0026ndash;8 weeks of gestation and identified the anterior lip cervical modulus and strain ratio as independent risk factors for preterm birth, constructing an early-pregnancy predictive model with favorable diagnostic performance. Similarly, Miao et al. [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] used E-cervix elastography to evaluate cervical tissue in 120 women with threatened preterm labor at 20\u0026ndash;32 weeks of gestation, finding that the proportion of hard cervical tissue was significantly lower in the preterm group than in the term group [(35.75\u0026thinsp;\u0026plusmn;\u0026thinsp;8.94)% vs. (61.30\u0026thinsp;\u0026plusmn;\u0026thinsp;10.69)%, F\u0026thinsp;=\u0026thinsp;156.88], and lower hard tissue proportion was moderately correlated with cervical shortening. These findings support the hypothesis that microstructural changes in the cervix (e.g., collagen remodeling, water content, and proteoglycan concentration) may precede macroscopic structural alterations [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAlthough ultrasound elastography shows promise as a noninvasive diagnostic tool, heterogeneity in instruments, technical approaches, patient selection, and study design limits its clinical application. In the present study, mid-gestation 2D ultrasound images were acquired using multiple devices and probes, and radiomic analysis was employed to extract deep cervical microstructural features. These features may quantitatively reflect cervical remodeling during pregnancy and its progression, demonstrating close association with preterm birth risk and providing a more precise quantitative basis for preterm birth prediction.\u003c/p\u003e\u003cp\u003eIn 1996, Iams et al. [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e] first demonstrated in a multicenter prospective study that a short cervix measured by transvaginal ultrasound serves as an independent predictor of preterm birth. Subsequent studies have reported that funneling of the internal cervical os often occurs prior to preterm birth [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. The histological characteristics of the internal cervical os differ significantly from those of the external os: while the external os is primarily composed of collagen fibers, the internal os contains approximately 50\u0026ndash;60% smooth muscle. These smooth muscle bundles encircle the internal os, maintaining cervical closure during pregnancy and exhibiting enhanced contractile capacity under oxytocin stimulation. Some researchers have proposed that the internal os possesses a specialized sphincter-like structure, which effectively ensures fetal retention in utero during gestation, and initiates labor only when its functional threshold is exceeded [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. This concept provides a clearer understanding of the role of the cervix in maintaining pregnancy and explains how cervical insufficiency can lead to preterm birth.\u003c/p\u003e\u003cp\u003eIn a study investigating the microstructure of the cervical internal os in pregnant rats [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], significant rearrangement of collagen fibers was observed in the internal os region during gestation, whereas other cervical regions showed minimal changes. As pregnancy progressed, collagen fibers in the internal os gradually shifted from an orientation parallel to the cervical canal to a vertical arrangement, which may influence the degree of internal os dilation and its biomechanical properties. Furthermore, when collagen fibers were parallel to the cervical canal, the radial displacement of the anterior and posterior lips of the cervix increased approximately 6.5-fold compared with the vertical orientation, indicating a direct relationship between collagen fiber remodeling and funneling of the internal os.\u003c/p\u003e\u003cp\u003eTherefore, focusing on the internal cervical os can enhance understanding of the mechanisms maintaining pregnancy and the pathophysiology of preterm birth, providing more precise predictive indicators and potential intervention strategies. In the present study, the combined radiomic features of the anterior (ROI1) and posterior (ROI2) lips of the internal os were used to construct the ROI1\u0026thinsp;+\u0026thinsp;ROI2 combined model, which demonstrated significant diagnostic performance in predicting preterm birth. This finding further supports the critical role of microstructural changes in the internal cervical os in both pregnancy maintenance and preterm birth prediction.\u003c/p\u003e\u003cp\u003eIn this study, a predictive model for preterm birth was established by integrating cervical ultrasound radiomic features with clinical parameters. The results demonstrated that a history of preterm birth and cervical length\u0026thinsp;\u0026le;\u0026thinsp;25 mm were independent risk factors, consistent with previous reports [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Cervical length, as a commonly used predictor, was further validated for its effectiveness in assessing preterm birth risk. Nine radiomic features were extracted from ultrasound images, including various high-order features reflecting tissue texture and gray-level distribution. Among them, wavelet-LL_glrlm_GrayLevelNonUniformity and log-sigma-3-0-mm-3D_glrlm_GrayLevelNonUniformity showed the strongest associations with preterm birth; these features were derived from the internal and external cervical os regions, reflecting gray-level heterogeneity and subtle textural differences, respectively. Except for the original shape feature (original_shape_MinorAxisLength), all other features were high-order, capturing microstructural differences that are difficult to detect with conventional visual assessment, indicating the potential value of radiomic features for noninvasive, early prediction.\u003c/p\u003e\u003cp\u003eFurthermore, decision curve analysis revealed that the combined model could substantially improve clinical benefit in practice, supporting its potential applicability. Our findings suggest that cervical ultrasound radiomic features can significantly enhance the prediction of preterm birth. By integrating multidimensional radiomic features with clinical data, the constructed model not only improves risk identification but also addresses the limitations of traditional single-parameter assessment.\u003c/p\u003e"},{"header":"Limitations","content":"\u003cp\u003eThis study is limited by its retrospective design and relatively small sample size, which may affect generalizability. Future prospective, multicenter, and larger-scale studies are warranted. Other factors such as maternal health and environmental influences should also be explored. This model requires validation in diverse populations and multicenter cohorts to confirm broader applicability.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eCervical ultrasound radiomics combined with clinical features effectively predicts preterm birth. This model improves early risk identification and supports timely clinical intervention, offering a promising tool for personalized management strategies.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003ePTB \u0026mdash; Preterm Birth\u003c/p\u003e\n\u003cp\u003eCL \u0026mdash; Cervical Length\u003c/p\u003e\n\u003cp\u003eROI \u0026mdash; Region of Interest\u003c/p\u003e\n\u003cp\u003eROI1 \u0026mdash; Cervical Internal Os Anterior Lip Region of Interest\u003c/p\u003e\n\u003cp\u003eROI2 \u0026mdash; Cervical Internal Os Posterior Lip Region of Interest\u003c/p\u003e\n\u003cp\u003eRad-score \u0026mdash; Radiomics Score\u003c/p\u003e\n\u003cp\u003eGLRLM \u0026mdash; Gray Level Run Length Matrix\u003c/p\u003e\n\u003cp\u003eROC \u0026mdash; Receiver Operating Characteristic\u003c/p\u003e\n\u003cp\u003eAUC \u0026mdash; Area Under the Curve\u003c/p\u003e\n\u003cp\u003eDCA \u0026mdash; Decision Curve Analysis\u003c/p\u003e\n\u003cp\u003eSVM \u0026mdash; Support Vector Machine\u003c/p\u003e\n\u003cp\u003eLR \u0026mdash; Logistic Regression\u003c/p\u003e\n\u003cp\u003eBMI \u0026mdash; Body Mass Index\u003c/p\u003e\n\u003cp\u003eSD \u0026mdash; Standard Deviation\u003c/p\u003e\n\u003cp\u003eDICOM \u0026mdash; Digital Imaging and Communications in Medicine\u003c/p\u003e\n\u003cp\u003eICC \u0026mdash; Intraclass Correlation Coefficient\u003c/p\u003e\n\u003cp\u003eESC \u0026mdash; European Society of Cardiology\u003c/p\u003e\n\u003cp\u003eSPSS \u0026mdash; Statistical Package for the Social Sciences\u003c/p\u003e\n\u003cp\u003eR \u0026mdash; R Statistical Software\u003c/p\u003e\n\u003cp\u003eOR \u0026mdash; Odds Ratio\u003c/p\u003e\n\u003cp\u003eCI \u0026mdash; Confidence Interval\u003c/p\u003e\n\u003cp\u003eUS \u0026mdash; Ultrasound\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eEthics approval and consent to participate\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Ethics Committee of Yichang Central People's Hospital, China Three Gorges University (approval number: 2024-11-14). Written informed consent was obtained from all participants prior to enrollment.\u003c/p\u003e\n\u003cp\u003eConsent for publication\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003eAvailability of data and materials\u003c/p\u003e\n\u003cp\u003eThe datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003eCompeting interests\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eNo funding was received for this study.\u003c/p\u003e\n\u003cp\u003eAuthors' contributions\u003c/p\u003e\n\u003cp\u003eW.H. and C.Z. conceived and designed the study. W.H. collected and organized the clinical and imaging data. H.S. and L.X. performed the radiomics feature extraction and statistical analyses. X.L. contributed to the interpretation of results and drafting of the manuscript. W.H. wrote the initial draft, and C.Z. critically revised the manuscript for important intellectual content. All authors reviewed and approved the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003eAcknowledgements\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank the staff of the Department of Ultrasound, The First College of Clinical Medical Science, China Three Gorges University \u0026amp; Yichang Central People's Hospital, for their support in data collection and technical assistance.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eVogel J P, Chawanpaiboon S, Moller A, et al. The global epidemiology of preterm birth[J]. Best Pract Res Clin Obstet Gynaecol. 2018,52:3-12.\u003c/li\u003e\n\u003cli\u003eWU P L，XUE Q，LIU X X，SUN X. Research progress on cervical cerclage for preventing preterm birth in twin pregnancies. Chinese Journal of Perinatal Medicine, 2024,27(3):258-261.(in Chinese).\u003c/li\u003e\n\u003cli\u003eZHONG H Y X H. Clinical guidelines for the prevention and treatment of preterm birth (version 2024). Chinese Journal of Obstetrics and Gynecology, 2024,59(04):257-269.(in Chinese).\u003c/li\u003e\n\u003cli\u003eXin W, Rixin S, Linrui L, Zhihui Q, Long L, Yu Z. Machine learning-based radiomics for predicting outcomes in cervical cancer patients undergoing concurrent chemoradiotherapy. Comput Biol Med. 2024;177:108593. \u003c/li\u003e\n\u003cli\u003eQiu H, Wang M, Wang S, et al. Integrating MRI-based radiomics and clinicopathological features for preoperative prognostication of early-stage cervical adenocarcinoma patients: in comparison to deep learning approach. Cancer Imaging. 2024;24(1):101.\u003c/li\u003e\n\u003cli\u003eThe Lancet. The unfinished agenda of preterm births[J]. Lancet. 2016;388(10058):2323. \u003c/li\u003e\n\u003cli\u003eSaccone G, Maruotti G M, Morlando M, et al. Randomized trial of screening for preterm birth in low-risk women - the preterm birth screening study[J]. Int J Nurs Stud , 2024,6(5):101267.\u003c/li\u003e\n\u003cli\u003eXU G C，ZHANG J P. Controversy and views on cervical cerclage. Chinese Journal of Perinatal Medicine, 2023,26(10):793-796.(in Chinese).\u003c/li\u003e\n\u003cli\u003eMoghaddam A O, Lin Z, Sivaguru M, et al. Heterogeneous microstructural changes of the cervix influence cervical funneling[J]. Acta Biomaterialia, 2022,140:434-445.\u003c/li\u003e\n\u003cli\u003eStone J, House M. Measurement of cervical softness before cerclage placement with an aspiration-based device[J]. Am J Obstet Gynecol MFM , 2023,5(4):100881.\u003c/li\u003e\n\u003cli\u003eSIHAN C，BING H U，XIAOZHEN X，LING D，YING Y. Construction of Prediction Model for Preterm Birth in Early Pregnancy via Cervical Elastography. Chinese Journal of Medical Imaging, 2023,31(12):1298-1303.(in Chinese).\u003c/li\u003e\n\u003cli\u003eMOU Y T，LONG Y，LUO Y C，ZHANG S. Application of E-cervix elastography technology in pregnant women with threatened preterm birth: an analysis of 120 cases. Chinese Journal of Perinatal Medicine, 2024,27(08):656-661.(in Chinese). \u003c/li\u003e\n\u003cli\u003ePatberg ET, Wells M, Vahanian SA, et al. Use of cervical elastography at 18 to 22 weeks\u0026apos; gestation in the prediction of spontaneous preterm birth[J]. Am J Obstet Gynecol. 2021;225(5):525.e1-525.e9.\u003c/li\u003e\n\u003cli\u003eIams J D, Goldenberg R L, Meis P J, et al. The length of the cervix and the risk of spontaneous premature delivery. National Institute of Child Health and Human Development Maternal Fetal Medicine Unit Network[J]. N Engl J Med, 1996,334(9):567-572.\u003c/li\u003e\n\u003cli\u003eCoutinho C M, Sotiriadis A, Odibo A, et al. ISUOG Practice Guidelines: role of ultrasound in the prediction of spontaneous preterm birth[J]. Ultrasound Obstet Gynecol. 2022,60(3):435-456.\u003c/li\u003e\n\u003cli\u003eMarinescu P S, Young R C, Miller L A, et al. Mid-trimester uterine electromyography in patients with a short cervix[J]. Am J Obstet Gynecol, 2022,227(1):81-83.\u003c/li\u003e\n\u003cli\u003eMoghaddam A O, Lin Z, Sivaguru M, et al. Heterogeneous microstructural changes of the cervix influence cervical funneling[J]. Acta Biomaterialia, 2022,140:434-445.\u003c/li\u003e\n\u003cli\u003eHessami K, D\u0026apos;Alberti E, Mascio D D, et al. Universal cervical length screening and risk of spontaneous preterm birth: a systematic review and meta-analysis[J]. Am J Obstet Gynecol MFM, 2024,6(5S):101343.\u003c/li\u003e\n\u003cli\u003eHessami K, D\u0026apos;Alberti E, Mascio D D, et al. Universal cervical length screening and risk of spontaneous preterm birth: a systematic review and meta-analysis[J]. Am J Obstet Gynecol MFM, 2024,6(5S):101343.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"bmc-pregnancy-and-childbirth","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"prch","sideBox":"Learn more about [BMC Pregnancy and Childbirth](http://bmcpregnancychildbirth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/prch/default.aspx","title":"BMC Pregnancy and Childbirth","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Ultrasound, Radiomics, Preterm birth, Cervix, Machine learning","lastPublishedDoi":"10.21203/rs.3.rs-7568405/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7568405/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eObjective: To evaluate the feasibility of predicting preterm birth using an ultrasound radiomics model combined with clinical features.\u003c/p\u003e\u003cp\u003eMethods: We retrospectively analyzed 521 pregnant women who underwent prenatal care at Yichang Central People\u0026rsquo;s Hospital between January 2018 and August 2024. Patients were randomly assigned to a training set (n\u0026thinsp;=\u0026thinsp;417) and a validation set (n\u0026thinsp;=\u0026thinsp;104) at an 8:2 ratio. Radiomic features were extracted from the region of interest (ROI) of the cervix on 2D ultrasound images, and combined with clinical high-risk factors. All features were standardized and normalized to a (0,1) range. Feature selection was performed using variance thresholding, optimal feature selection (by number and percentage), and significance-based filtering. Logistic regression, random forest, and support vector machine models were constructed for preterm birth prediction. Model performance and clinical utility were evaluated using ROC curves, AUC, Hosmer-Lemeshow test, and decision curve analysis (DCA).\u003c/p\u003e\u003cp\u003eResults: The combined model achieved the highest predictive performance for preterm birth, with AUCs of 0.874 and 0.841 in the training and validation sets, respectively, indicating good consistency. Hosmer-Lemeshow test and decision curves demonstrated good model calibration and high clinical net benefit.\u003c/p\u003e\u003cp\u003eConclusion: Ultrasound radiomics combined with clinical features can effectively predict preterm birth and may support early, non-invasive clinical interventions.\u003c/p\u003e","manuscriptTitle":"Prediction of Preterm Birth Based on Cervical Ultrasound Radiomics Combined with Clinical Features","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-17 12:42:22","doi":"10.21203/rs.3.rs-7568405/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2025-10-24T08:51:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"337744384783732831180651359319707849055","date":"2025-10-14T07:02:08+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-10-06T21:26:06+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-09-11T15:01:55+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-09-10T09:10:43+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-09-10T09:10:21+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Pregnancy and Childbirth","date":"2025-09-09T02:08:56+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-pregnancy-and-childbirth","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"prch","sideBox":"Learn more about [BMC Pregnancy and Childbirth](http://bmcpregnancychildbirth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/prch/default.aspx","title":"BMC Pregnancy and Childbirth","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"466a964a-665c-4254-8c4b-2647a17ce0dd","owner":[],"postedDate":"October 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2025-10-17T12:42:22+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-17 12:42:22","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7568405","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7568405","identity":"rs-7568405","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}