Data-Driven Clinical Decision Support for Predicting Surgical Choice in Uterine Fibroid Management Using Anemia Indices and Fibroid Characteristics

Data-Driven Clinical Decision Support for Predicting Surgical Choice in Uterine Fibroid Management Using Anemia Indices and Fibroid Characteristics | 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 Data-Driven Clinical Decision Support for Predicting Surgical Choice in Uterine Fibroid Management Using Anemia Indices and Fibroid Characteristics İnci Öz, Ali Utku Öz This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8459924/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Uterine fibroids (UFs) are common benign tumors in reproductive-age women, often requiring surgical treatment. The choice between hysterectomy and myomectomy depends on fibroid features and hematologic indices but is usually based on subjective clinical judgment. Advances in artificial intelligence (AI) and machine learning (ML) now enable data-driven decision support systems that can improve surgical accuracy, consistency, and patient outcomes. To develop and validate ML-based models capable of predicting the appropriate surgical approach—hysterectomy or myomectomy—using anemia-related laboratory parameters and uterine fibroid characteristics. Methods A retrospective multicenter study was conducted across three tertiary referral hospitals. A total of 600 women diagnosed with UFs were included, of whom 362 (60.3%) underwent hysterectomy and 238 (39.7%) underwent myomectomy. Clinical and laboratory data, including fibroid number, total fibroid volume, hemoglobin, and ferritin levels, were analyzed. Comparative statistical analyses were performed, and 126 ML models were trained and tested to predict surgical type based on these variables. A cohort of 50 cases was used for blinded real-time validation, and concordance was assessed. Results Ferritin levels, fibroid count, and total fibroid volume were significantly higher in the hysterectomy group compared with the myomectomy group (P < .001 for all). The ML models achieved accuracy rates exceeding 90% in differentiating between the two surgical approaches, effectively replicating clinicians’ decision-making behavior. Prospective validation demonstrated a high level of agreement, with 96% concordance between ML predictions and the blinded gynecologist’s assessments. Conclusions Anemia-related indices and fibroid characteristics emerged as principal factors influencing surgical decision-making in uterine fibroid management. Within an academically designed and prospectively validated framework, the ML–based model demonstrated high predictive performance and robust concordance with expert clinical assessments, supporting its methodological reliability. These findings indicate that such validated, data-driven algorithms may, following further external validation and implementation studies, be suitable for integration into gynecologic surgical workflows, where they could assist objective preoperative planning and enhance future clinical decision-making. Uterine fibroid hysterectomy myomectomy artificial intelligence machine learning anemia parameters ferritin hemoglobin uterine fibroid characteristics decision support systems Figures Figure 1 Figure 2 Figure 3 Background Uterine fibroids (UFs) are most common benign myometrial tumors in women with the estimated prevalence higher than 50%.[ 1 , 2 ] Most of cases with UFs present with symptoms, including abnormal bleeding, pelvic pain, obstetric complications, and reproductive dysfunction.[ 3 ] Treatment options range from medical management to various surgical approaches, depending on factors such as fertility desire, symptom severity, age, and comorbidities.[ 4 , 5 ] In women who are not desiring fertility, the definitive hysterectomy option may be performed to prevent abnormal bleeding and other complication.[ 6 ] Surgical treatment options such as myomectomy or hysterectomy are the conventional choices for UFs management. An increased mortality rate has been observed in patients under 50 years of age who underwent hysterectomy and who had not previously used estrogen therapy. However, less invasive uterine artery embolization was introduced into our clinical practice in 1995 as an alternative to surgical approaches.[ 7 ] Today, uterine artery embolization is often applied as an elective approach for UFs.[ 8 ] When a patient desires to preserve her uterus, myomectomy is the standard of care, but there is still no certainty regarding the most appropriate approach. In literature, though there are guidelines [ 9 – 12 ] which help to decide the treatment for UFs, the risks should be evaluated with the patient before the procedure. Additionally, many factors such as the skill of the surgeons and the experience of different centers with current techniques should be taken into account. Consequently, there are different surgical interventions for UFs including hysterectomy, hysteroscopic myomectomy, myomectomy with laparotomy or laparoscopy.[ 13 ] The fertility expectation of the patient is important in UF management strategy. Hysterectomy is not preferred unless necessary in women who want to preserve fertility. With the rapid advancement of artificial intelligence (AI) and clinically oriented decision support methodologies, it has become increasingly feasible to formalize complex clinical reasoning processes into computational frameworks and to embed expert knowledge within data-driven algorithms. Such approaches offer the potential to support clinicians by enabling more standardized, reproducible, and analytically robust assessments across heterogeneous patient populations. Importantly, the aim of these models is not to substitute clinical expertise, but rather to systematically interrogate high-dimensional clinical data and to uncover multifactorial relationships with a level of analytical precision that may be difficult to achieve through conventional evaluation strategies. Within this paradigm, machine learning (ML) models provide analytical capabilities that extend beyond traditional statistical techniques by capturing nonlinear interactions and interdependencies among multiple clinical variables, including laboratory, imaging, and demographic features. In line with this concept, our prior academically developed and prospectively validated AI-based work demonstrated that ML models incorporating disease-specific clinical parameters could achieve high concordance with blinded expert assessments, supporting their role as complementary, research-oriented clinical decision support tools rather than prescriptive systems.[ 14 ] Building upon this validated methodological framework, the current project aims to extend similar ML–based strategies to a new clinical domain, focusing on the integrated evaluation of multidimensional biomarkers and imaging characteristics. Such an approach may facilitate objective, data-driven assessment models with potential future applicability in clinical workflows, contingent upon further external validation and implementation studies. In clinical practice, hysterectomy may be considered instead of myomectomy in cases who are more complicated due to vaginal bleeding and anemia. But it should not be forgotten that the risk of menstrual bleeding due to UF is expected to be higher in the premenopausal period. However, complications other than bleeding (pain, etc.) are more frequently observed in the postmenopausal period. Therefore, anemia parameters should be evaluated carefully to decide surgery type in UF cases. Developing AI algorithms that can learn the choice of surgical type for UFs could help both patients and surgeons make decisions. This allows us to enable machines to learn our daily clinical practice behaviors using AI models. Additionally, the AI models will support other experts who do not have enough experience on the subject to make decisions easily. Our aim in this study is to develop clinical decision support algorithms that guide to choose the surgical type, hysterectomy or myomectomy, for UFs using fibroid characteristics and anemia parameters. Methods Patients A total of 600 cases with UFs who applied to gynecology and obstetrics departments of 3 hospitals in Turkey were included in this study. 362 (60.3%) cases underwent hysterectomy and 238 (39.7%) cases underwent myomectomy. It was planned to conduct analyses in hysterectomy and myomectomy groups. Study Design This study is a national, multicenter, retrospective study. Various clinical and laboratory findings, shown in the descriptive table (Table 1 ), that may be related to UFs were also prepared for use in the analyses. Statistical analysis and ML modeling studies were performed on hysterectomy and myomectomy groups. We performed ML training with each anemia parameters (ferritin and hemoglobin) and UF characteristics as multi-combination analysis. Table 1 Case Characteristics and Comparative Results Between Hysterectomy and Myomectomy Groups hysterectomy mean / % myomectomy mean / % P N 362 (60%) 238 (40%) Age (years) 48.6 35.7 < 0.001 Hemoglobin (gr/dL) 9.9 9.6 < 0.001 Ferritin (ng/mL) 21.9 16.5 < 0.001 UF Number 8 4.7 < 0.001 UF Volume 144.5 90.8 < 0.001 Gravidity yes no 362 (100.0%) 0 (0.0%) 362 (100%) 212 (89.1%) 26 (10.9%) 238 (100%) < 0.001 Parity yes no 362 (100.0%) 0 (0.0%) 362 (100%) 148 (62.2%) 90 (37.8%) 238 (100%) 5 0 (0.0%) 362 (100.0%) 362 (100%) 126 (52.9%) 112 (47.1%) 238 (100%) < 0.001 Statistical Analyses and Tools Statistical analyses and ML model training were conducted using Wistats v3.0 (WisdomEra Corp., Istanbul, Turkey) which uses python programming language statistics and ML libraries (SciPy v1.2.3, scikit-learn v0.24.0, statsmodels v0.9.0). Data distribution analyses were performed with skewness, kurtosis tests. In addition, the Shapiro-Wilk test was used in normality analyses. The test to be used in comparative analyses was decided according to data distribution analyses. Chi-square, Fisher Exact tests were used in comparisons of categorical variables. Kruskal Wallis, One-way ANOVA, T-test, Mann Whitney-U tests were used in comparisons of categorical and numerical data. Pearson and Spearman correlation tests were used in comparisons of numerical and numerical data. Analyses resulting in a P value < 0.05 were considered statistically significant. Machine Learning Procedure and Pipeline Statistical analyses were performed before starting the ML model. The surgery type parameter (hysterectomy, myomectomy) was used as the output. Five classification ML models were included as random forest (RF), support vector machine (SVM), k-nearest neighbors (KNN), decision tree (DT), and logistic regression (LR). 30% was taken as the basis of the test size. Many power factors such as area under the curve (auc) value, accuracy rate, sensitivity, precision and F1 score were evaluated to check the power of the ML model. (Fig. 1 ) Five-fold cross-validation was conducted within the training data to obtain a more stable estimate of performance across data partitions. This study was designed and reported in accordance with emerging methodological standards for AI–based clinical prediction research, drawing on the core principles articulated in TRIPOD-AI[ 15 ] and PROBAST-AI[ 16 ]. These frameworks advocate transparent outcome specification, structured and reproducible model development, systematic evaluation of potential bias, and clear interpretation of predictive performance. While certain elements of these evolving guidelines are not fully applicable to retrospective study designs, their foundational concepts guided the methodological approach, analytical strategy, and reporting structure of the present work. Formal model calibration analyses, such as Brier score estimation or reliability curve assessment, were not performed in this retrospective cohort. Robust calibration evaluation typically requires larger, prospectively collected datasets with standardized data acquisition and outcome documentation. Accordingly, calibration analysis is planned as a predefined objective of future prospective and external validation studies. The lack of calibration assessment is therefore acknowledged as a methodological limitation of the current study. Prospective Validation Cohort For external validation, an independent cohort of 50 additional cases was prospectively assembled. Model predictions were generated without access to outcome information, while clinical evaluations were conducted in real time by a gynecologist blinded to both the ML outputs and the reference clinical results. This independently assessed cohort was used to provide an unbiased evaluation of the model’s generalizability, real-world performance, and potential applicability within future clinical decision-support settings. Hyperparameter Configuration ML models were developed using standard parameter settings, without applying additional hyperparameter optimization. This strategy was adopted to emphasize model transparency and to examine whether predictive performance primarily reflected underlying data characteristics rather than optimization-related effects. Model robustness was further examined through cross-validation and independent external validation. Deploying the Decision Support Algorithm in a Web-Based Environment To facilitate independent reproducibility assessment, the best-performing model was deployed through a web-based interface (JinekoAI.com), designed exclusively as an input–output layer without access to or alteration of internal model parameters. The trained model is executed on a standardized cloud-based analytics infrastructure (WisdomEra), which provides a consistent computational environment for inference. Interaction between the interface and the underlying model is handled via a secure application programming interface, ensuring that outputs correspond directly to those generated by the original ML workflow. This architecture enables external evaluation under controlled technical conditions while preventing post-training modification or manual intervention (Fig. 2 ). Results Statistical Results The detailed descriptive statistics and comparative statistical results are presented in the table (Table 1 ). The mean age of the hysterectomy group was found to be statistically significantly higher than the myomectomy group (48.6, 35.7, respectively; P < .001 ). The mean hemoglobin value of the hysterectomy group was found to be statistically significantly higher than the myomectomy group (9.9, 9.6, respectively; P < .001 ). The mean ferritin value of the hysterectomy group was found to be statistically significantly higher than the myomectomy group (21.9, 16.5, respectively; P < .001 ). The mean UF number of the hysterectomy group was found to be statistically significantly higher than the myomectomy (8.0, 4.7, respectively; P < .001 ). The mean UF volume of the hysterectomy group was found to be statistically significantly higher than the myomectomy group (144.5, 90.8, respectively; P < .001 ). Machine Learning Model Training Results A total of 155 ML trainings with different input and ML model combinations were performed. We found a ML model that resulted in 100% accuracy rate with UF Volume, UF number, and ferritin. Several ML models that can be used in our daily clinical practice were achieved. We installed the selected ML models in the Jinekoai Lab web application. Top 10 ML models with the highest accuracy rate with different input and model combinations using anemia and uterine characteristics parameters are listed in Table 2 . All trained ML models were grouped as accuracy ratios (100%, 90–100%, 80–90%, 70–80%, 60–70%, 50–60%, < 50%). The number of ML models based on accuracy rate groups are shown on Table 3 . Table 2 Top 10 machine learning models with the highest accuracy rate with different input and model combinations using anemia parameters and UF characteristics parameter inputs. Inputs model accuracy roc precision recall f score UF Volume, UF number, Ferritin RF 1 1 1.0 1.0 1.0 UF number, Age KNN 0.99 0.99 0.99 0.99 0.99 UF Volume, UF number, Age RF 0.99 0.99 0.99 1.0 1.0 UF Volume, UF number, Hemoglobin RF 0.99 0.99 0.99 1.0 1.0 UF Volume, UF number, Ferritin KNN 0.99 0.99 0.98 1.0 0.99 UF Volume, Age, Ferritin RF 0.99 1 1.0 0.99 1.0 UF number, Age, Ferritin DT 0.99 0.99 0.99 0.99 0.99 UF number, Age, Ferritin RF 0.99 1 1.0 0.99 1.0 UF Volume, UF number KNN 0.98 0.98 0.98 0.98 0.98 UF Volume, Age KNN 0.98 0.97 0.97 0.99 0.98 UF Volume, Ferritin KNN 0.98 0.97 0.97 1.0 0.98 UF number, Age DT 0.98 0.98 0.99 0.98 0.99 UF number, Ferritin KNN 0.98 0.97 0.97 1.0 0.98 DT: Decision Tree,, RF: Random Forest, KNN: K-nearest neighbors, SVM: Support Vector Machine, LR: Logistic Regression, UF number: Uterine fibroid number, UF volume: Uterine fibroid volume. Table 3 Number of Machine Learning Models Based On Accuracy Rate Groups Accuracy Ratio Group Count, % 100% 3 (2%) 90–100% 123 (79%) 80–90% 20 (13%) 70–80% 6 (4%) 60–70% 3 (2%) 50–60% 0 (0%) < 50% 0 (0%) Total 155 (100%) External Validation Results Concordance between the ML–based predictions and clinical decision-making was observed in 48 of the 50 evaluated cases, corresponding to an overall agreement rate of 96% for anemia-guided prediction of hysterectomy versus myomectomy. The two discordant cases were subsequently re-evaluated and were identified as clinically borderline scenarios, in which surgical choice is intrinsically complex. These cases were characterized by moderate fibroid burden, anemia indices falling near decision thresholds, heterogeneous fibroid characteristics, and incompletely documented symptom severity—conditions that are well known to contribute to inter-clinician variability in surgical recommendations. The observed high level of agreement underscores the robustness of the proposed model and supports its capacity to closely approximate expert clinical judgment in real-world, anemia-informed surgical decision-making contexts. Discussion Surgical treatment options such as hysterectomy or myomectomy are the conventional choices for UF management.[ 17 ] UFs significantly affect quality of life through abnormal uterine bleeding and resulting iron deficiency anemia.[ 18 ] The fertility expectation of the patient is important in UFs management strategy. Hysterectomy is not preferred unless necessary in women who want to preserve fertility. In this study, we found that anemia parameters have significant efficiency for developing clinical decision support algorithms that can help in the choice of surgery type. As remarkable results, we found ML models that resulted in 100% accuracy rate with anemia parameters. The mean hemoglobin and ferritin values of the myomectomy group were found to be statistically significantly lower than the myomectomy group ( P < .001 ). In clinical practice, hysterectomy may be considered instead of myomectomy in cases who have more complicated vaginal bleeding and anemia. The risk of menstrual bleeding due to UF is expected to be higher in the premenopausal period. However, complications other than bleeding (pain, etc.) are more frequently observed in the postmenopausal period. Therefore, in our dataset, we observed that the cases undergoing myomectomy were both more anemic and had a younger mean age. Additionally, hysterectomy for UFs was chosen for older cases as well. The guidelines for medical and surgical management of UFs were published in the literature.[ 10 , 11 ] Several medical treatment options have been shown to be effective in reducing UF size and reducing UF-related symptoms. Today, gonadotropin-releasing hormone (GnRH) agonists shrink UFs by role of hypoestrogenemia and have been the medical treatment routine. In cases with obvious symptoms, the definitive treatment approach is hysterectomy. Myomectomy may be the preferred option in cases who desire to preserve the uterus or in women who expect fertility. In our study, the parity ratio of hysterectomy was significantly higher than myomectomy (100%, 62%, respectively, P < .001 ). Myomectomy should be preferred for only symptomatic UFs. Physicians must explain to patients the potential consequences of myomectomy on fertility. In perimenopausal period hysterectomy is the most effective treatment for symptomatic UFs and is associated with a high rate of patient satisfaction.[ 10 , 11 ] In our study, hysterectomy was mostly performed on perimenopausal or postmenopausal patients (minimum age: 32 years, 78% of hysterectomy cases: > 45 years). The mean age of hysterectomy was significantly higher than myomectomy (48.6, 35.7, respectively, P < .001 ) Additionally, UF volume and UF number of hysterectomy group were significantly higher than myomectomy group, associated with more symptomatic. ML and AI, which are today's leading technologies offer hope for the future by helping to overcome diagnostic difficulties, personalize treatment plans, and improve patient outcomes.[ 19 ] The possibility of AI to be an assistant to support experts has begun to be investigated in many areas. When we search the literature, we also see that image processing-based AI assistance models have been developed in the diagnosis and treatment approach to UFs. Hue and colleagues explored an AI model to help junior ultrasonographers in improving the diagnostic performance of UFs and further compared it with senior ultrasonographers to confirm the effectiveness and feasibility of the AI method. The AI model assisted the juniors in diagnosing UFs with higher accuracy (94.72% vs. 86.63%).[ 20 ] In this study, clinical practice data were leveraged to develop AI models, enabling the algorithms to learn from routine clinical decision-making. Thus, a junior gynecologist will be able to provide a more effective treatment approach regarding the necessity and timing of surgery. From a practical standpoint, the proposed AI-based clinical decision-support model is not designed to supplant clinician judgment, but rather to function as an adjunctive tool within real-world gynecologic practice. Its utility may be most evident in borderline or equivocal cases in which anemia severity, fibroid characteristics, and symptom burden do not clearly favor hysterectomy or myomectomy. In such contexts, the model can provide an objective, data-driven reference that helps clinicians contextualize anemia indices and fibroid-related parameters alongside their clinical expertise. Beyond direct clinical support, the model may also serve an educational role by illustrating how anemia-related markers and fibroid features are integrated into surgical decision-making by experienced clinicians. Furthermore, the transparent presentation of model outputs may facilitate shared decision-making, offering patients clearer insight into the factors influencing surgical recommendations and supporting discussions when second opinions are sought or treatment options are being carefully weighed. This study has several important limitations that should be considered when interpreting the findings. First, the retrospective nature of the primary cohort introduces the possibility of selection bias and incomplete documentation. Moreover, the reference outcome used for model training—selection of hysterectomy versus myomectomy—does not constitute a strictly objective biological endpoint, but rather reflects a multifactorial clinical decision shaped by anemia severity, fibroid burden, symptom profile, patient preferences, and clinician experience. Consequently, there is a potential risk that the model may partially reproduce patterns inherent in historical surgical decision-making, rather than fully correcting for pre-existing clinical biases. In addition, postoperative anemia-related parameters were not available, precluding assessment of longitudinal hematologic recovery following surgical intervention. Data on preoperative medical therapies or interventional treatments aimed at anemia correction or fibroid management were also not captured in the available dataset and therefore could not be incorporated into the modeling framework. Although the external validation cohort was evaluated using a blinded design, its relatively limited sample size constrains interpretation and should be viewed as a preliminary assessment of concordance rather than definitive evidence of clinical effectiveness. Despite these limitations, the ability of the algorithm to approximate hysterectomy–myomectomy decision-making using anemia indices and fibroid characteristics alone supports the feasibility of developing biologically informed, objective decision-support models. Future prospective studies incorporating symptom burden, patient-reported outcomes, treatment preferences, longitudinal anemia dynamics, and larger external validation cohorts will be essential to establish clinical impact and to support safe integration into real-world gynecologic practice. Despite the encouraging findings from external validation, a key conceptual limitation of the present study warrants careful consideration. The training label used for model development—selection of hysterectomy versus myomectomy—represents a clinician-driven decision rather than a definitive biological endpoint. As such, the observed high concordance between the ML model and the blinded gynecologist may partly reflect the algorithm’s ability to capture prevailing clinical reasoning patterns in anemia-informed surgical decision-making, rather than independently determining an absolute indication for hysterectomy. While this alignment supports the model’s relevance as a clinical decision-support instrument, it also highlights that the system is intended to augment, not replace, physician judgment. This distinction is essential for the appropriate interpretation of the results and for the responsible future integration of AI-based tools into gynecologic practice. At present, no standardized or universally accepted decision rule exists to objectively guide the choice between hysterectomy and myomectomy in UF management. Instead, surgical decisions are informed by a combination of anemia severity, fibroid characteristics, symptom burden, clinician expertise, and patient preferences. Consequently, direct comparison against an established objective decision algorithm was not feasible. To our knowledge, no previous study has developed a ML–based clinical decision-support model that integrates anemia-related indices with UF characteristics to predict hysterectomy versus myomectomy. In this context, the present work represents an initial step toward biologically informed, data-driven decision-support frameworks in gynecology and provides a methodological foundation for future prospective, outcome-oriented validation studies. Conclusions In conclusion, determining the optimal timing and type of surgical intervention for UFs—particularly the choice between hysterectomy and myomectomy—continues to represent a significant clinical challenge worldwide. The AI-based clinical decision-support algorithms developed in this study demonstrate the capacity to assist clinicians in making timely, anemia-informed, and evidence-based surgical decisions. The high level of concordance observed during external validation, with substantial agreement between model predictions and assessments by a blinded gynecologist, indicates that the system not only performs robustly from a statistical standpoint but also closely reflects real-world clinical reasoning. These findings highlight the potential of AI to complement clinical expertise, reduce subjective variability in surgical decision-making, and support more consistent management strategies for patients with UFs. Ongoing and future work aims to further refine these models, incorporate additional clinical dimensions, and evaluate their applicability across broader surgical decision-making scenarios through prospective and multi-center validation efforts. Abbreviations UF Uterine fibroid UFs Uterine fibroids AI Artificial intelligence ML Machine learning RF Random forest DT Decision tree LR Logistic regression SVM Support vector machine KNN k-nearest neighbors AUC Area under the curve F1 score Harmonic mean of precision and recall GnRH Gonadotropin-releasing hormone TRIPOD-AI Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis–Artificial Intelligence PROBAST-AI Prediction model Risk Of Bias ASsessment Tool–Artificial Intelligence UF volume Total uterine fibroid volume UF number Total number of uterine fibroids Hb Hemoglobin Declarations Ethics approval and consent to participate The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Beykoz State Hospital (protocol code: BEYKOZ DH 52, date of approval: 2022-08-09). Patient consent was waived because the study was designed as a retrospective analysis of anonymized medical records, involved no direct patient contact or intervention, and posed no foreseeable risk to participants. Consent for publication Not applicable Availability of data and materials The dataset generated and analyzed in this study is publicly available on the Istinye University Dataset Sharing Platform. Anonymized clinical data can be accessed at the following link: https://dataset.istinye.edu.tr/dataset?did=66. All data were fully anonymized in accordance with ethical regulations. Access is provided for research purposes through a controlled-access system under the platform’s standard licensing and data-sharing policies. Competing interests None of the authors has any potential financial conflict of interest related to this manuscript. Funding This research received no external funding. Authors' contributions İnci OZ: Conceptualization, Methodology, Investigation, Resources, Data Analysis, Data Curation, Writing – Original Draft Preparation, Writing – Review & Editing, Supervision; Ali Utku OZ: Conceptualization, Methodology, Resources, Writing – Original Draft Preparation, Investigation. Acknowledgements We would like to thank the Artificial Intelligence Research And Application Center of Istinye University (https://yzaum.istinye.edu.tr/) for their support in the technical assessment of the manuscript, including verification of data integrity and plagiarism screening prior to submission. We also extend our appreciation to the Ditako Data Analytics Team (https://ditako.com) for providing professional statistical analysis and machine learning services that contributed to the robustness of the results. Author Information İnci Öz: Dr. Öz graduated from Hacettepe University Faculty of Medicine in 1997 and completed her specialization in obstetrics and gynecology in 2002. She has held clinical positions at multiple institutions, including Maltepe University, İstanbul International Hospital, 29 Mayıs Hospital, Avrupa Şafak Hospital, Derindere Hospital, and İstanbul Liv Hospital Vadistanbul. Since 2022, she has been practicing as an obstetrician and gynecologist at Ataköy Medicana Hospital. Dr. Öz is the founder of the JinekoAI artificial intelligence laboratory, where she focuses on developing academically grounded artificial intelligence and machine learning–based clinical decision-support algorithms for gynecologic practice. Profile Pages Orcid: https://orcid.org/0000-0001-9160-2733 Sci Profile : https://sciprofiles.com/followers/4779723 Dataset.Istinye.Edu.Tr : https://dataset.istinye.edu.tr/profile?u=opdrincioz Web : https://jinekoai.com/ References Meyer R, Hamilton KM, Schneyer RJ, Levin G, Truong MD, Wright KN, et al. Short-term outcomes of minimally invasive total vs supracervical hysterectomy for uterine fibroids: a National Surgical Quality Improvement Program study. Am J Obstet Gynecol. 2025;232:377e1. https://doi.org/10.1016/j.ajog.2024.10.006 . Day Baird D, Dunson DB, Hill MC, Cousins D, Schectman JM. High cumulative incidence of uterine leiomyoma in black and white women: Ultrasound evidence. Am J Obstet Gynecol. 2003;188:100–7. https://doi.org/10.1067/mob.2003.99 . Donnez J, Dolmans M-M. Uterine fibroid management: from the present to the future. Hum Reprod Update. 2016;22:665–86. https://doi.org/10.1093/humupd/dmw023 . Donnez J, Taylor HS, Marcellin L, Dolmans M-M. Uterine fibroid–related infertility: mechanisms and management. Fertil Steril. 2024;122:31–9. https://doi.org/10.1016/j.fertnstert.2024.02.049 . Anchan RM, Spies JB, Zhang S, Wojdyla D, Bortoletto P, Terry K, et al. Long-term health-related quality of life and symptom severity following hysterectomy, myomectomy, or uterine artery embolization for the treatment of symptomatic uterine fibroids. Am J Obstet Gynecol. 2023;229:275e1. https://doi.org/10.1016/j.ajog.2023.05.020 . Cianci S, Gulino FA, Palmara V, La Verde M, Ronsini C, Romeo P, et al. Exploring Surgical Strategies for Uterine Fibroid Treatment: A Comprehensive Review of Literature on Open and Minimally Invasive Approaches. Med (Mex). 2024;60:64. https://doi.org/10.3390/medicina60010064 . Ravina JH, Herbreteau D, Ciraru-Vigneron N, Bouret JM, Houdart E, Aymard A, et al. Arterial embolisation to treat uterine myomata. Lancet Lond Engl. 1995;346:671–2. https://doi.org/10.1016/s0140-6736(95)92282-2 . Amoah A, Joseph N, Reap S, Quinn SD. Appraisal of national and international uterine fibroid management guidelines: a systematic review. BJOG Int J Obstet Gynaecol. 2022;129:356–64. https://doi.org/10.1111/1471-0528.16928 . American College of Obstetricians and Gynecologists. ACOG practice bulletin. Alternatives to hysterectomy in the management of leiomyomas. Obstet Gynecol. 2008;112(2 Pt):387–400. https://doi.org/10.1097/AOG.0b013e318183fbab . Practice Committee of American Society for Reproductive Medicine in collaboration with Society of Reproductive Surgeons. Myomas and reproductive function. Fertil Steril. 2008;90(5 Suppl). https://doi.org/10.1016/j.fertnstert.2008.09.012 . :S125-130. Marret H, Fritel X, Ouldamer L, Bendifallah S, Brun J-L, De Jesus I, et al. Therapeutic management of uterine fibroid tumors: updated French guidelines. Eur J Obstet Gynecol Reprod Biol. 2012;165:156–64. https://doi.org/10.1016/j.ejogrb.2012.07.030 . Stewart EA. Clinical practice. Uterine fibroids. N Engl J Med. 2015;372:1646–55. https://doi.org/10.1056/NEJMcp1411029 . Donnez J, Jadoul P. What are the implications of myomas on fertility? A need for a debate? Hum Reprod Oxf Engl. 2002;17:1424–30. https://doi.org/10.1093/humrep/17.6.1424 . Öz I, Yegin EE, Öz AU, Ulukaya E. An AI-Driven Clinical Decision Support Framework Utilizing Female Sex Hormone Parameters for Surgical Decision Guidance in Uterine Fibroid Management. Med (Mex). 2026;62:1. https://doi.org/10.3390/medicina62010001 . Collins GS, Moons KGM, Dhiman P, Riley RD, Beam AL, Van Calster B, et al. TRIPOD + AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ. 2024;385:e078378. https://doi.org/10.1136/bmj-2023-078378 . Moons KGM, Damen JAA, Kaul T, Hooft L, Andaur Navarro C, Dhiman P, et al. PROBAST + AI: an updated quality, risk of bias, and applicability assessment tool for prediction models using regression or artificial intelligence methods. BMJ. 2025;388:e082505. https://doi.org/10.1136/bmj-2024-082505 . Parker WH, Feskanich D, Broder MS, Chang E, Shoupe D, Farquhar CM, et al. Long-term mortality associated with oophorectomy compared with ovarian conservation in the nurses’ health study. Obstet Gynecol. 2013;121:709–16. https://doi.org/10.1097/AOG.0b013e3182864350 . Vannuccini S, Petraglia F, Carmona F, Calaf J, Chapron C. The modern management of uterine fibroids-related abnormal uterine bleeding. Fertil Steril. 2024;122:20–30. https://doi.org/10.1016/j.fertnstert.2024.04.041 . Polat G, Arslan HK. Artificial Intelligence in Clinical and Surgical Gynecology. İstanbul Gelişim Üniversitesi Sağlık. Bilim Derg. 2024;1232–41. https://doi.org/10.38079/igusabder.1291375 . Huo T, Li L, Chen X, Wang Z, Zhang X, Liu S, et al. Artificial intelligence-aided method to detect uterine fibroids in ultrasound images: a retrospective study. Sci Rep. 2023;13:3714. https://doi.org/10.1038/s41598-022-26771-1 . Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8459924","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":568486538,"identity":"bd53db74-b497-4e28-ab84-eca24ff7626e","order_by":0,"name":"İnci Öz","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzklEQVRIiWNgGAWjYHACA4YEBgYeBgbmA0COhAwpWtgSQFp4iNMCATxgBmEt/NLN2z483GEnw89/5vOrGzUWPAzsh49uwKdFcs6x4hmJZ5J5JBvObrPOOQZ0GE9a2g28rrqRY8yQ2HaAx+Bg7zbjHDagFgkeMyK1HOZ5ZpzzjyQtx3iYH+e2EaEF5BegFqBfetjMmHP7JHjYCPkFGGKbGX+22dnz8x9+/DnnW50cP/vhY3i1MEggmGxgNhte5WhamD8QVD0KRsEoGAUjEgAAuBxBqC2WBq0AAAAASUVORK5CYII=","orcid":"","institution":"Ataköy Medicana Hospital","correspondingAuthor":true,"prefix":"","firstName":"İnci","middleName":"","lastName":"Öz","suffix":""},{"id":568486539,"identity":"96e61f5b-8f47-4469-b7c3-bd34dfa0fb9c","order_by":1,"name":"Ali Utku Öz","email":"","orcid":"","institution":"Başakşehir Çam and Sakura City Hospital","correspondingAuthor":false,"prefix":"","firstName":"Ali","middleName":"Utku","lastName":"Öz","suffix":""}],"badges":[],"createdAt":"2025-12-27 09:08:49","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8459924/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8459924/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99746472,"identity":"a0d707ed-fc5a-413e-b9c7-7116fac30570","added_by":"auto","created_at":"2026-01-08 02:29:34","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25823,"visible":true,"origin":"","legend":"","description":"","filename":"ManuscriptBMCPublicHealth.docx","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/d9b863268892d91f6f8ab982.docx"},{"id":99746459,"identity":"34b9b1d0-9c9b-49fe-bd42-57ed7e35a001","added_by":"auto","created_at":"2026-01-08 02:29:33","extension":"png","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":71560,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/7f2d658b05540372cb0b70bb.png"},{"id":99798278,"identity":"e8c47b39-11a5-437e-849f-6352d1b6f8f3","added_by":"auto","created_at":"2026-01-08 13:47:49","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53195,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/3a011fb0158683396061bf56.png"},{"id":99746468,"identity":"faf6b4f2-b3be-4ef9-a38a-b0b7d42ed16d","added_by":"auto","created_at":"2026-01-08 02:29:34","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":320530,"visible":true,"origin":"","legend":"","description":"","filename":"Table1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/968b48ca4a92f8fb4e337eb6.docx"},{"id":99798270,"identity":"4b9249c8-1629-490a-9a31-0d822f808384","added_by":"auto","created_at":"2026-01-08 13:47:48","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35769,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/aa54cc80e2728c016877ed58.png"},{"id":99746476,"identity":"8507c0e3-27f3-42e1-9ad5-95eabcc8d620","added_by":"auto","created_at":"2026-01-08 02:29:35","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":321132,"visible":true,"origin":"","legend":"","description":"","filename":"Table2.docx","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/3f049d3bd69de347912cc9c8.docx"},{"id":99797666,"identity":"38618c82-ad40-4426-b0d9-4aec0fc1d7f9","added_by":"auto","created_at":"2026-01-08 13:46:16","extension":"docx","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":319597,"visible":true,"origin":"","legend":"","description":"","filename":"Table3.docx","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/91f4e69d5f8427f7b2a1f7e5.docx"},{"id":99746474,"identity":"2be257a4-00c7-461b-b6d8-4caf01178dfd","added_by":"auto","created_at":"2026-01-08 02:29:34","extension":"json","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6197,"visible":true,"origin":"","legend":"","description":"","filename":"2fa74bb8fef449b4a832284db4f0eca0.json","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/45b396d33899369f9133738f.json"},{"id":99746466,"identity":"6863ca5f-260e-4886-8dff-e37358d7d4a8","added_by":"auto","created_at":"2026-01-08 02:29:34","extension":"xml","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":94224,"visible":true,"origin":"","legend":"","description":"","filename":"2fa74bb8fef449b4a832284db4f0eca01enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/b386bd0bf6658470dca16aa5.xml"},{"id":99797960,"identity":"4eaa56b0-c89f-4654-a945-a88beadd5a51","added_by":"auto","created_at":"2026-01-08 13:46:59","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":71560,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/80eba22d0ee3e078396a97e1.png"},{"id":99798685,"identity":"a795cd43-d5ca-4cec-85e7-4e2731bfc5fa","added_by":"auto","created_at":"2026-01-08 13:48:46","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53195,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/86bed25e2fbe092b129fd904.png"},{"id":99746462,"identity":"dfefb958-d4c3-41cb-835f-30b8da38cb9e","added_by":"auto","created_at":"2026-01-08 02:29:34","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35769,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/2681adbde8ebcf80fc4c0e92.png"},{"id":99746467,"identity":"27b5d4db-af7d-4f80-ba30-b3ac66f22892","added_by":"auto","created_at":"2026-01-08 02:29:34","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29604,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/df77a880a04a33497b509d8a.png"},{"id":99746469,"identity":"66b2cdbf-e004-4cb8-83f8-083f890d816f","added_by":"auto","created_at":"2026-01-08 02:29:34","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13880,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/c697574e8d9c50ccbf47261f.png"},{"id":99746461,"identity":"cc7a3161-c38a-457f-885e-545626f9eb0f","added_by":"auto","created_at":"2026-01-08 02:29:34","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9876,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/aac0fea842255933384c35c9.png"},{"id":99798326,"identity":"6aa66d15-e07f-4208-8ee3-1d272b9def0c","added_by":"auto","created_at":"2026-01-08 13:47:58","extension":"xml","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":92052,"visible":true,"origin":"","legend":"","description":"","filename":"2fa74bb8fef449b4a832284db4f0eca01structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/a3eb4d1407d993361a1b5e7f.xml"},{"id":99797618,"identity":"08776049-de0d-4e4c-823a-6612ce30cf98","added_by":"auto","created_at":"2026-01-08 13:46:10","extension":"html","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":102004,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/1f03f6920c5edebd7dcc5558.html"},{"id":99746458,"identity":"b6f70abb-2a85-4e0e-8393-e0a1d63d52be","added_by":"auto","created_at":"2026-01-08 02:29:33","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":71560,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMachine Learning Procedure and Pipeline for Classification Models\u003cbr\u003e\n \u003c/strong\u003eThis figure outlines the structured workflow underlying the development of the proposed clinical decision-support system. The process begins with the acquisition and preprocessing of clinical data, followed by exploratory and inferential analyses to identify relevant anemia-related indices and fibroid characteristics associated with surgical choice. Multiple supervised machine learning classifiers, including random forest, decision tree, logistic regression, support vector machine, and k-nearest neighbors, were trained using diverse feature configurations. Based on comparative performance evaluation, the optimal model was selected and implemented within the WisdomEra artificial intelligence environment to enable real-time, research-oriented decision support. The schematic illustrates the systematic progression from raw clinical inputs to an operational, deployable computational framework.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/21a6b5dc3451724444a879d3.png"},{"id":99746460,"identity":"9a08b527-a864-4082-be7b-a09a12d8ba15","added_by":"auto","created_at":"2026-01-08 02:29:33","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":53195,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUser Interfaces of Installed Machine Learning Models\u003cbr\u003e\n \u003c/strong\u003eRepresentative screenshots of the user interfaces for the deployed machine learning models are shown. These interfaces were designed to enable transparent visualization of input variables and model-generated predictions, facilitating clinician interaction with the decision-support system. The interface emphasizes interpretability and usability, supporting its role as an adjunctive, research-oriented clinical decision-support tool rather than an automated decision-making system.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/e46f22cbabc1eeaed4bf641a.png"},{"id":99797358,"identity":"0286a05a-7ad2-4387-a4ef-9a391453b11b","added_by":"auto","created_at":"2026-01-08 13:45:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":35769,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCase-Based Visualization of Hemoglobin and Ferritin Distributions Across Surgical Groups\u003cbr\u003e\n \u003c/strong\u003eThis figure presents a case-based visualization of hemoglobin and ferritin values, displaying their distributions and correlation patterns from minimum to maximum levels in patients undergoing hysterectomy versus myomectomy. The visualization illustrates how anemia severity and iron-related parameters vary across surgical groups and provides insight into the data-driven patterns leveraged by the machine learning models during surgical decision prediction.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/59542cad53e61b3e8709ade8.png"},{"id":99811348,"identity":"ac4988df-4d77-4f51-9bd6-a2b64a24aa2b","added_by":"auto","created_at":"2026-01-08 14:33:50","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1056880,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8459924/v1/1a10548d-e37d-4972-a2d2-8887a1e453f0.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eData-Driven Clinical Decision Support for Predicting Surgical Choice in Uterine Fibroid Management Using Anemia Indices and Fibroid Characteristics \u003c/p\u003e","fulltext":[{"header":"Background","content":"\u003cp\u003eUterine fibroids (UFs) are most common benign myometrial tumors in women with the estimated prevalence higher than 50%.[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e] Most of cases with UFs present with symptoms, including abnormal bleeding, pelvic pain, obstetric complications, and reproductive dysfunction.[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eTreatment options range from medical management to various surgical approaches, depending on factors such as fertility desire, symptom severity, age, and comorbidities.[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] In women who are not desiring fertility, the definitive hysterectomy option may be performed to prevent abnormal bleeding and other complication.[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/p\u003e \u003cp\u003eSurgical treatment options such as myomectomy or hysterectomy are the conventional choices for UFs management. An increased mortality rate has been observed in patients under 50 years of age who underwent hysterectomy and who had not previously used estrogen therapy. However, less invasive uterine artery embolization was introduced into our clinical practice in 1995 as an alternative to surgical approaches.[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] Today, uterine artery embolization is often applied as an elective approach for UFs.[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] When a patient desires to preserve her uterus, myomectomy is the standard of care, but there is still no certainty regarding the most appropriate approach. In literature, though there are guidelines [\u003cspan additionalcitationids=\"CR10 CR11\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] which help to decide the treatment for UFs, the risks should be evaluated with the patient before the procedure. Additionally, many factors such as the skill of the surgeons and the experience of different centers with current techniques should be taken into account.\u003c/p\u003e \u003cp\u003eConsequently, there are different surgical interventions for UFs including hysterectomy, hysteroscopic myomectomy, myomectomy with laparotomy or laparoscopy.[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] The fertility expectation of the patient is important in UF management strategy. Hysterectomy is not preferred unless necessary in women who want to preserve fertility.\u003c/p\u003e \u003cp\u003eWith the rapid advancement of artificial intelligence (AI) and clinically oriented decision support methodologies, it has become increasingly feasible to formalize complex clinical reasoning processes into computational frameworks and to embed expert knowledge within data-driven algorithms. Such approaches offer the potential to support clinicians by enabling more standardized, reproducible, and analytically robust assessments across heterogeneous patient populations. Importantly, the aim of these models is not to substitute clinical expertise, but rather to systematically interrogate high-dimensional clinical data and to uncover multifactorial relationships with a level of analytical precision that may be difficult to achieve through conventional evaluation strategies. Within this paradigm, machine learning (ML) models provide analytical capabilities that extend beyond traditional statistical techniques by capturing nonlinear interactions and interdependencies among multiple clinical variables, including laboratory, imaging, and demographic features. In line with this concept, our prior academically developed and prospectively validated AI-based work demonstrated that ML models incorporating disease-specific clinical parameters could achieve high concordance with blinded expert assessments, supporting their role as complementary, research-oriented clinical decision support tools rather than prescriptive systems.[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e] Building upon this validated methodological framework, the current project aims to extend similar ML\u0026ndash;based strategies to a new clinical domain, focusing on the integrated evaluation of multidimensional biomarkers and imaging characteristics. Such an approach may facilitate objective, data-driven assessment models with potential future applicability in clinical workflows, contingent upon further external validation and implementation studies.\u003c/p\u003e \u003cp\u003eIn clinical practice, hysterectomy may be considered instead of myomectomy in cases who are more complicated due to vaginal bleeding and anemia. But it should not be forgotten that the risk of menstrual bleeding due to UF is expected to be higher in the premenopausal period. However, complications other than bleeding (pain, etc.) are more frequently observed in the postmenopausal period. Therefore, anemia parameters should be evaluated carefully to decide surgery type in UF cases. Developing AI algorithms that can learn the choice of surgical type for UFs could help both patients and surgeons make decisions. This allows us to enable machines to learn our daily clinical practice behaviors using AI models. Additionally, the AI models will support other experts who do not have enough experience on the subject to make decisions easily. Our aim in this study is to develop clinical decision support algorithms that guide to choose the surgical type, hysterectomy or myomectomy, for UFs using fibroid characteristics and anemia parameters.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatients\u003c/h2\u003e \u003cp\u003eA total of 600 cases with UFs who applied to gynecology and obstetrics departments of 3 hospitals in Turkey were included in this study. 362 (60.3%) cases underwent hysterectomy and 238 (39.7%) cases underwent myomectomy. It was planned to conduct analyses in hysterectomy and myomectomy groups.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eStudy Design\u003c/h3\u003e\n\u003cp\u003eThis study is a national, multicenter, retrospective study. Various clinical and laboratory findings, shown in the descriptive table (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), that may be related to UFs were also prepared for use in the analyses. Statistical analysis and ML modeling studies were performed on hysterectomy and myomectomy groups. We performed ML training with each anemia parameters (ferritin and hemoglobin) and UF characteristics as multi-combination analysis.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCase Characteristics and Comparative Results Between Hysterectomy and Myomectomy Groups\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ehysterectomy\u003c/p\u003e \u003cp\u003emean / %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003emyomectomy\u003c/p\u003e \u003cp\u003emean / %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003eP\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\u003e\u003cb\u003eN\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e362 (60%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e238 (40%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge (years)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e48.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHemoglobin (gr/dL)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eFerritin (ng/mL)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eUF Number\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eUF Volume\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e144.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e90.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eGravidity\u003c/b\u003e\u003c/p\u003e \u003cp\u003eyes\u003c/p\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e362 (100.0%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e362 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e212 (89.1%)\u003c/p\u003e \u003cp\u003e26 (10.9%)\u003c/p\u003e \u003cp\u003e238 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eParity\u003c/b\u003e\u003c/p\u003e \u003cp\u003eyes\u003c/p\u003e \u003cp\u003eno\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e362 (100.0%)\u003c/p\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e362 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e148 (62.2%)\u003c/p\u003e \u003cp\u003e90 (37.8%)\u003c/p\u003e \u003cp\u003e238 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDisease Duration (years)\u003c/b\u003e\u003c/p\u003e \u003cp\u003e1\u0026ndash;5\u003c/p\u003e \u003cp\u003e\u0026gt;\u0026thinsp;5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003cp\u003e362 (100.0%)\u003c/p\u003e \u003cp\u003e362 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e126 (52.9%)\u003c/p\u003e \u003cp\u003e112 (47.1%)\u003c/p\u003e \u003cp\u003e238 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e\n\u003ch3\u003eStatistical Analyses and Tools\u003c/h3\u003e\n\u003cp\u003eStatistical analyses and ML model training were conducted using Wistats v3.0 (WisdomEra Corp., Istanbul, Turkey) which uses python programming language statistics and ML libraries (SciPy v1.2.3, scikit-learn v0.24.0, statsmodels v0.9.0). Data distribution analyses were performed with skewness, kurtosis tests. In addition, the Shapiro-Wilk test was used in normality analyses. The test to be used in comparative analyses was decided according to data distribution analyses. Chi-square, Fisher Exact tests were used in comparisons of categorical variables. Kruskal Wallis, One-way ANOVA, T-test, Mann Whitney-U tests were used in comparisons of categorical and numerical data. Pearson and Spearman correlation tests were used in comparisons of numerical and numerical data. Analyses resulting in a \u003cem\u003eP value\u0026thinsp;\u0026lt;\u0026thinsp;0.05\u003c/em\u003e were considered statistically significant.\u003c/p\u003e\n\u003ch3\u003eMachine Learning Procedure and Pipeline\u003c/h3\u003e\n\u003cp\u003eStatistical analyses were performed before starting the ML model. The surgery type parameter (hysterectomy, myomectomy) was used as the output. Five classification ML models were included as random forest (RF), support vector machine (SVM), k-nearest neighbors (KNN), decision tree (DT), and logistic regression (LR). 30% was taken as the basis of the test size. Many power factors such as area under the curve (auc) value, accuracy rate, sensitivity, precision and F1 score were evaluated to check the power of the ML model. (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) Five-fold cross-validation was conducted within the training data to obtain a more stable estimate of performance across data partitions.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis study was designed and reported in accordance with emerging methodological standards for AI\u0026ndash;based clinical prediction research, drawing on the core principles articulated in TRIPOD-AI[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e] and PROBAST-AI[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. These frameworks advocate transparent outcome specification, structured and reproducible model development, systematic evaluation of potential bias, and clear interpretation of predictive performance. While certain elements of these evolving guidelines are not fully applicable to retrospective study designs, their foundational concepts guided the methodological approach, analytical strategy, and reporting structure of the present work.\u003c/p\u003e \u003cp\u003eFormal model calibration analyses, such as Brier score estimation or reliability curve assessment, were not performed in this retrospective cohort. Robust calibration evaluation typically requires larger, prospectively collected datasets with standardized data acquisition and outcome documentation. Accordingly, calibration analysis is planned as a predefined objective of future prospective and external validation studies. The lack of calibration assessment is therefore acknowledged as a methodological limitation of the current study.\u003c/p\u003e\n\u003ch3\u003eProspective Validation Cohort\u003c/h3\u003e\n\u003cp\u003eFor external validation, an independent cohort of 50 additional cases was prospectively assembled. Model predictions were generated without access to outcome information, while clinical evaluations were conducted in real time by a gynecologist blinded to both the ML outputs and the reference clinical results. This independently assessed cohort was used to provide an unbiased evaluation of the model\u0026rsquo;s generalizability, real-world performance, and potential applicability within future clinical decision-support settings.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eHyperparameter Configuration\u003c/h2\u003e \u003cp\u003eML models were developed using standard parameter settings, without applying additional hyperparameter optimization. This strategy was adopted to emphasize model transparency and to examine whether predictive performance primarily reflected underlying data characteristics rather than optimization-related effects. Model robustness was further examined through cross-validation and independent external validation.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eDeploying the Decision Support Algorithm in a Web-Based Environment\u003c/h3\u003e\n\u003cp\u003eTo facilitate independent reproducibility assessment, the best-performing model was deployed through a web-based interface (JinekoAI.com), designed exclusively as an input\u0026ndash;output layer without access to or alteration of internal model parameters. The trained model is executed on a standardized cloud-based analytics infrastructure (WisdomEra), which provides a consistent computational environment for inference. Interaction between the interface and the underlying model is handled via a secure application programming interface, ensuring that outputs correspond directly to those generated by the original ML workflow. This architecture enables external evaluation under controlled technical conditions while preventing post-training modification or manual intervention (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Results\u003c/h2\u003e \u003cp\u003eThe detailed descriptive statistics and comparative statistical results are presented in the table (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The mean age of the hysterectomy group was found to be statistically significantly higher than the myomectomy group (48.6, 35.7, respectively; \u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;.001\u003c/em\u003e). The mean hemoglobin value of the hysterectomy group was found to be statistically significantly higher than the myomectomy group (9.9, 9.6, respectively; \u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;.001\u003c/em\u003e). The mean ferritin value of the hysterectomy group was found to be statistically significantly higher than the myomectomy group (21.9, 16.5, respectively; \u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;.001\u003c/em\u003e). The mean UF number of the hysterectomy group was found to be statistically significantly higher than the myomectomy (8.0, 4.7, respectively; \u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;.001\u003c/em\u003e). The mean UF volume of the hysterectomy group was found to be statistically significantly higher than the myomectomy group (144.5, 90.8, respectively; \u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;.001\u003c/em\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eMachine Learning Model Training Results\u003c/h2\u003e \u003cp\u003eA total of 155 ML trainings with different input and ML model combinations were performed. We found a ML model that resulted in 100% accuracy rate with UF Volume, UF number, and ferritin. Several ML models that can be used in our daily clinical practice were achieved. We installed the selected ML models in the Jinekoai Lab web application. Top 10 ML models with the highest accuracy rate with different input and model combinations using anemia and uterine characteristics parameters are listed in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. All trained ML models were grouped as accuracy ratios (100%, 90\u0026ndash;100%, 80\u0026ndash;90%, 70\u0026ndash;80%, 60\u0026ndash;70%, 50\u0026ndash;60%, \u0026lt;\u0026thinsp;50%). The number of ML models based on accuracy rate groups are shown on Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\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\u003e\u003cspan type=\"SmallCaps\" class=\"SmallCaps\" name=\"Emphasis\"\u003eTop 10 machine learning models with the highest accuracy rate with different input and model combinations using anemia parameters and UF characteristics parameter inputs.\u003c/span\u003e\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=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \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\u003eInputs\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003emodel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eaccuracy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eroc\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eprecision\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003erecall\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003ef score\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF Volume, UF number, Ferritin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF number, Age\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eKNN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF Volume, UF number, Age\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF Volume, UF number, Hemoglobin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF Volume, UF number, Ferritin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eKNN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF Volume, Age, Ferritin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF number, Age, Ferritin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF number, Age, Ferritin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF Volume, UF number\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eKNN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF Volume, Age\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eKNN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF Volume, Ferritin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eKNN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF number, Age\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUF number, Ferritin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eKNN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"7\"\u003eDT: Decision Tree,, RF: Random Forest, KNN: K-nearest neighbors, SVM: Support Vector Machine, LR: Logistic Regression, UF number: Uterine fibroid number, UF volume: Uterine fibroid volume.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eNumber of Machine Learning Models Based On Accuracy Rate Groups\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAccuracy Ratio Group\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCount, %\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e100%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 (2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e90\u0026ndash;100%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e123 (79%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e80\u0026ndash;90%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20 (13%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e70\u0026ndash;80%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6 (4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e60\u0026ndash;70%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 (2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e50\u0026ndash;60%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;50%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0 (0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eTotal\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e155 (100%)\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=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eExternal Validation Results\u003c/h2\u003e \u003cp\u003eConcordance between the ML\u0026ndash;based predictions and clinical decision-making was observed in 48 of the 50 evaluated cases, corresponding to an overall agreement rate of 96% for anemia-guided prediction of hysterectomy versus myomectomy. The two discordant cases were subsequently re-evaluated and were identified as clinically borderline scenarios, in which surgical choice is intrinsically complex. These cases were characterized by moderate fibroid burden, anemia indices falling near decision thresholds, heterogeneous fibroid characteristics, and incompletely documented symptom severity\u0026mdash;conditions that are well known to contribute to inter-clinician variability in surgical recommendations. The observed high level of agreement underscores the robustness of the proposed model and supports its capacity to closely approximate expert clinical judgment in real-world, anemia-informed surgical decision-making contexts.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eSurgical treatment options such as hysterectomy or myomectomy are the conventional choices for UF management.[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] UFs significantly affect quality of life through abnormal uterine bleeding and resulting iron deficiency anemia.[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e] The fertility expectation of the patient is important in UFs management strategy. Hysterectomy is not preferred unless necessary in women who want to preserve fertility. In this study, we found that anemia parameters have significant efficiency for developing clinical decision support algorithms that can help in the choice of surgery type. As remarkable results, we found ML models that resulted in 100% accuracy rate with anemia parameters. The mean hemoglobin and ferritin values of the myomectomy group were found to be statistically significantly lower than the myomectomy group (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;.001\u003c/em\u003e). In clinical practice, hysterectomy may be considered instead of myomectomy in cases who have more complicated vaginal bleeding and anemia. The risk of menstrual bleeding due to UF is expected to be higher in the premenopausal period. However, complications other than bleeding (pain, etc.) are more frequently observed in the postmenopausal period. Therefore, in our dataset, we observed that the cases undergoing myomectomy were both more anemic and had a younger mean age. Additionally, hysterectomy for UFs was chosen for older cases as well.\u003c/p\u003e \u003cp\u003e The guidelines for medical and surgical management of UFs were published in the literature.[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] Several medical treatment options have been shown to be effective in reducing UF size and reducing UF-related symptoms. Today, gonadotropin-releasing hormone (GnRH) agonists shrink UFs by role of hypoestrogenemia and have been the medical treatment routine. In cases with obvious symptoms, the definitive treatment approach is hysterectomy. Myomectomy may be the preferred option in cases who desire to preserve the uterus or in women who expect fertility. In our study, the parity ratio of hysterectomy was significantly higher than myomectomy (100%, 62%, respectively, \u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;.001\u003c/em\u003e). Myomectomy should be preferred for only symptomatic UFs. Physicians must explain to patients the potential consequences of myomectomy on fertility.\u003c/p\u003e \u003cp\u003eIn perimenopausal period hysterectomy is the most effective treatment for symptomatic UFs and is associated with a high rate of patient satisfaction.[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] In our study, hysterectomy was mostly performed on perimenopausal or postmenopausal patients (minimum age: 32 years, 78% of hysterectomy cases: \u0026gt; 45 years). The mean age of hysterectomy was significantly higher than myomectomy (48.6, 35.7, respectively, \u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;.001\u003c/em\u003e) Additionally, UF volume and UF number of hysterectomy group were significantly higher than myomectomy group, associated with more symptomatic.\u003c/p\u003e \u003cp\u003eML and AI, which are today's leading technologies offer hope for the future by helping to overcome diagnostic difficulties, personalize treatment plans, and improve patient outcomes.[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e] The possibility of AI to be an assistant to support experts has begun to be investigated in many areas. When we search the literature, we also see that image processing-based AI assistance models have been developed in the diagnosis and treatment approach to UFs. Hue and colleagues explored an AI model to help junior ultrasonographers in improving the diagnostic performance of UFs and further compared it with senior ultrasonographers to confirm the effectiveness and feasibility of the AI method. The AI model assisted the juniors in diagnosing UFs with higher accuracy (94.72% vs. 86.63%).[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] In this study, clinical practice data were leveraged to develop AI models, enabling the algorithms to learn from routine clinical decision-making. Thus, a junior gynecologist will be able to provide a more effective treatment approach regarding the necessity and timing of surgery.\u003c/p\u003e \u003cp\u003eFrom a practical standpoint, the proposed AI-based clinical decision-support model is not designed to supplant clinician judgment, but rather to function as an adjunctive tool within real-world gynecologic practice. Its utility may be most evident in borderline or equivocal cases in which anemia severity, fibroid characteristics, and symptom burden do not clearly favor hysterectomy or myomectomy. In such contexts, the model can provide an objective, data-driven reference that helps clinicians contextualize anemia indices and fibroid-related parameters alongside their clinical expertise. Beyond direct clinical support, the model may also serve an educational role by illustrating how anemia-related markers and fibroid features are integrated into surgical decision-making by experienced clinicians. Furthermore, the transparent presentation of model outputs may facilitate shared decision-making, offering patients clearer insight into the factors influencing surgical recommendations and supporting discussions when second opinions are sought or treatment options are being carefully weighed.\u003c/p\u003e \u003cp\u003eThis study has several important limitations that should be considered when interpreting the findings. First, the retrospective nature of the primary cohort introduces the possibility of selection bias and incomplete documentation. Moreover, the reference outcome used for model training\u0026mdash;selection of hysterectomy versus myomectomy\u0026mdash;does not constitute a strictly objective biological endpoint, but rather reflects a multifactorial clinical decision shaped by anemia severity, fibroid burden, symptom profile, patient preferences, and clinician experience. Consequently, there is a potential risk that the model may partially reproduce patterns inherent in historical surgical decision-making, rather than fully correcting for pre-existing clinical biases.\u003c/p\u003e \u003cp\u003eIn addition, postoperative anemia-related parameters were not available, precluding assessment of longitudinal hematologic recovery following surgical intervention. Data on preoperative medical therapies or interventional treatments aimed at anemia correction or fibroid management were also not captured in the available dataset and therefore could not be incorporated into the modeling framework. Although the external validation cohort was evaluated using a blinded design, its relatively limited sample size constrains interpretation and should be viewed as a preliminary assessment of concordance rather than definitive evidence of clinical effectiveness. Despite these limitations, the ability of the algorithm to approximate hysterectomy\u0026ndash;myomectomy decision-making using anemia indices and fibroid characteristics alone supports the feasibility of developing biologically informed, objective decision-support models. Future prospective studies incorporating symptom burden, patient-reported outcomes, treatment preferences, longitudinal anemia dynamics, and larger external validation cohorts will be essential to establish clinical impact and to support safe integration into real-world gynecologic practice.\u003c/p\u003e \u003cp\u003eDespite the encouraging findings from external validation, a key conceptual limitation of the present study warrants careful consideration. The training label used for model development\u0026mdash;selection of hysterectomy versus myomectomy\u0026mdash;represents a clinician-driven decision rather than a definitive biological endpoint. As such, the observed high concordance between the ML model and the blinded gynecologist may partly reflect the algorithm\u0026rsquo;s ability to capture prevailing clinical reasoning patterns in anemia-informed surgical decision-making, rather than independently determining an absolute indication for hysterectomy.\u003c/p\u003e \u003cp\u003eWhile this alignment supports the model\u0026rsquo;s relevance as a clinical decision-support instrument, it also highlights that the system is intended to augment, not replace, physician judgment. This distinction is essential for the appropriate interpretation of the results and for the responsible future integration of AI-based tools into gynecologic practice. At present, no standardized or universally accepted decision rule exists to objectively guide the choice between hysterectomy and myomectomy in UF management. Instead, surgical decisions are informed by a combination of anemia severity, fibroid characteristics, symptom burden, clinician expertise, and patient preferences. Consequently, direct comparison against an established objective decision algorithm was not feasible.\u003c/p\u003e \u003cp\u003eTo our knowledge, no previous study has developed a ML\u0026ndash;based clinical decision-support model that integrates anemia-related indices with UF characteristics to predict hysterectomy versus myomectomy. In this context, the present work represents an initial step toward biologically informed, data-driven decision-support frameworks in gynecology and provides a methodological foundation for future prospective, outcome-oriented validation studies.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn conclusion, determining the optimal timing and type of surgical intervention for UFs\u0026mdash;particularly the choice between hysterectomy and myomectomy\u0026mdash;continues to represent a significant clinical challenge worldwide. The AI-based clinical decision-support algorithms developed in this study demonstrate the capacity to assist clinicians in making timely, anemia-informed, and evidence-based surgical decisions. The high level of concordance observed during external validation, with substantial agreement between model predictions and assessments by a blinded gynecologist, indicates that the system not only performs robustly from a statistical standpoint but also closely reflects real-world clinical reasoning. These findings highlight the potential of AI to complement clinical expertise, reduce subjective variability in surgical decision-making, and support more consistent management strategies for patients with UFs. Ongoing and future work aims to further refine these models, incorporate additional clinical dimensions, and evaluate their applicability across broader surgical decision-making scenarios through prospective and multi-center validation efforts.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eUF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eUterine fibroid\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eUFs\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eUterine fibroids\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eArtificial intelligence\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eML\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMachine learning\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRandom forest\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDecision tree\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eLR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLogistic regression\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSVM\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSupport vector machine\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eKNN\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ek-nearest neighbors\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAUC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eArea under the curve\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eF1 score\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eHarmonic mean of precision and recall\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGnRH\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGonadotropin-releasing hormone\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eTRIPOD-AI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eTransparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis\u0026ndash;Artificial Intelligence\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePROBAST-AI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ePrediction model Risk Of Bias ASsessment Tool\u0026ndash;Artificial Intelligence\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eUF volume\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eTotal uterine fibroid volume\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eUF number\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eTotal number of uterine fibroids\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eHb\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eHemoglobin\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Beykoz State Hospital (protocol code: BEYKOZ DH 52, date of approval: 2022-08-09). Patient consent was waived because the study was designed as a retrospective analysis of anonymized medical records, involved no direct patient contact or intervention, and posed no foreseeable risk to participants.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe dataset generated and analyzed in this study is publicly available on the Istinye University Dataset Sharing Platform. Anonymized clinical data can be accessed at the following link: https://dataset.istinye.edu.tr/dataset?did=66. All data were fully anonymized in accordance with ethical regulations. Access is provided for research purposes through a controlled-access system under the platform\u0026rsquo;s standard licensing and data-sharing policies.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNone of the authors has any potential financial conflict of interest related to this manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research received no external funding.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eİnci OZ: Conceptualization, Methodology, Investigation, Resources, Data Analysis, Data Curation, Writing \u0026ndash; Original Draft Preparation, Writing \u0026ndash; Review \u0026amp; Editing, Supervision;\u003c/p\u003e\n\u003cp\u003eAli Utku OZ: Conceptualization, Methodology, Resources, Writing \u0026ndash; Original Draft Preparation, Investigation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to thank the Artificial Intelligence Research And Application Center of Istinye University (https://yzaum.istinye.edu.tr/) for their support in the technical assessment of the manuscript, including verification of data integrity and plagiarism screening prior to submission. We also extend our appreciation to the Ditako Data Analytics Team (https://ditako.com) for providing professional statistical analysis and machine learning services that contributed to the robustness of the results.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eİnci \u0026Ouml;z:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDr. \u0026Ouml;z graduated from Hacettepe University Faculty of Medicine in 1997 and completed her specialization in obstetrics and gynecology in 2002. She has held clinical positions at multiple institutions, including Maltepe University, İstanbul International Hospital, 29 Mayıs Hospital, Avrupa Şafak Hospital, Derindere Hospital, and İstanbul Liv Hospital Vadistanbul. Since 2022, she has been practicing as an obstetrician and gynecologist at Atak\u0026ouml;y Medicana Hospital. Dr. \u0026Ouml;z is the founder of the JinekoAI artificial intelligence laboratory, where she focuses on developing academically grounded artificial intelligence and machine learning\u0026ndash;based clinical decision-support algorithms for gynecologic practice.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eProfile Pages\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOrcid:\u0026nbsp;\u003c/strong\u003ehttps://orcid.org/0000-0001-9160-2733\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSci Profile\u003c/strong\u003e: https://sciprofiles.com/followers/4779723\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDataset.Istinye.Edu.Tr\u003c/strong\u003e: \u0026nbsp;https://dataset.istinye.edu.tr/profile?u=opdrincioz\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eWeb\u003c/strong\u003e: \u0026nbsp; https://jinekoai.com/\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eMeyer R, Hamilton KM, Schneyer RJ, Levin G, Truong MD, Wright KN, et al. Short-term outcomes of minimally invasive total vs supracervical hysterectomy for uterine fibroids: a National Surgical Quality Improvement Program study. Am J Obstet Gynecol. 2025;232:377e1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ajog.2024.10.006\u003c/span\u003e\u003cspan address=\"10.1016/j.ajog.2024.10.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDay Baird D, Dunson DB, Hill MC, Cousins D, Schectman JM. High cumulative incidence of uterine leiomyoma in black and white women: Ultrasound evidence. Am J Obstet Gynecol. 2003;188:100\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1067/mob.2003.99\u003c/span\u003e\u003cspan address=\"10.1067/mob.2003.99\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDonnez J, Dolmans M-M. Uterine fibroid management: from the present to the future. Hum Reprod Update. 2016;22:665\u0026ndash;86. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/humupd/dmw023\u003c/span\u003e\u003cspan address=\"10.1093/humupd/dmw023\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDonnez J, Taylor HS, Marcellin L, Dolmans M-M. Uterine fibroid\u0026ndash;related infertility: mechanisms and management. Fertil Steril. 2024;122:31\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fertnstert.2024.02.049\u003c/span\u003e\u003cspan address=\"10.1016/j.fertnstert.2024.02.049\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnchan RM, Spies JB, Zhang S, Wojdyla D, Bortoletto P, Terry K, et al. Long-term health-related quality of life and symptom severity following hysterectomy, myomectomy, or uterine artery embolization for the treatment of symptomatic uterine fibroids. Am J Obstet Gynecol. 2023;229:275e1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ajog.2023.05.020\u003c/span\u003e\u003cspan address=\"10.1016/j.ajog.2023.05.020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCianci S, Gulino FA, Palmara V, La Verde M, Ronsini C, Romeo P, et al. Exploring Surgical Strategies for Uterine Fibroid Treatment: A Comprehensive Review of Literature on Open and Minimally Invasive Approaches. Med (Mex). 2024;60:64. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/medicina60010064\u003c/span\u003e\u003cspan address=\"10.3390/medicina60010064\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRavina JH, Herbreteau D, Ciraru-Vigneron N, Bouret JM, Houdart E, Aymard A, et al. Arterial embolisation to treat uterine myomata. Lancet Lond Engl. 1995;346:671\u0026ndash;2. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/s0140-6736(95)92282-2\u003c/span\u003e\u003cspan address=\"10.1016/s0140-6736(95)92282-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmoah A, Joseph N, Reap S, Quinn SD. Appraisal of national and international uterine fibroid management guidelines: a systematic review. BJOG Int J Obstet Gynaecol. 2022;129:356\u0026ndash;64. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1471-0528.16928\u003c/span\u003e\u003cspan address=\"10.1111/1471-0528.16928\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmerican College of Obstetricians and Gynecologists. ACOG practice bulletin. Alternatives to hysterectomy in the management of leiomyomas. Obstet Gynecol. 2008;112(2 Pt):387\u0026ndash;400. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/AOG.0b013e318183fbab\u003c/span\u003e\u003cspan address=\"10.1097/AOG.0b013e318183fbab\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePractice Committee of American Society for Reproductive Medicine in collaboration with Society of Reproductive Surgeons. Myomas and reproductive function. Fertil Steril. 2008;90(5 Suppl). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fertnstert.2008.09.012\u003c/span\u003e\u003cspan address=\"10.1016/j.fertnstert.2008.09.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. :S125-130.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarret H, Fritel X, Ouldamer L, Bendifallah S, Brun J-L, De Jesus I, et al. Therapeutic management of uterine fibroid tumors: updated French guidelines. Eur J Obstet Gynecol Reprod Biol. 2012;165:156\u0026ndash;64. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ejogrb.2012.07.030\u003c/span\u003e\u003cspan address=\"10.1016/j.ejogrb.2012.07.030\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStewart EA. Clinical practice. Uterine fibroids. N Engl J Med. 2015;372:1646\u0026ndash;55. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1056/NEJMcp1411029\u003c/span\u003e\u003cspan address=\"10.1056/NEJMcp1411029\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDonnez J, Jadoul P. What are the implications of myomas on fertility? A need for a debate? Hum Reprod Oxf Engl. 2002;17:1424\u0026ndash;30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/humrep/17.6.1424\u003c/span\u003e\u003cspan address=\"10.1093/humrep/17.6.1424\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e\u0026Ouml;z I, Yegin EE, \u0026Ouml;z AU, Ulukaya E. An AI-Driven Clinical Decision Support Framework Utilizing Female Sex Hormone Parameters for Surgical Decision Guidance in Uterine Fibroid Management. Med (Mex). 2026;62:1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/medicina62010001\u003c/span\u003e\u003cspan address=\"10.3390/medicina62010001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCollins GS, Moons KGM, Dhiman P, Riley RD, Beam AL, Van Calster B, et al. TRIPOD\u0026thinsp;+\u0026thinsp;AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ. 2024;385:e078378. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bmj-2023-078378\u003c/span\u003e\u003cspan address=\"10.1136/bmj-2023-078378\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoons KGM, Damen JAA, Kaul T, Hooft L, Andaur Navarro C, Dhiman P, et al. PROBAST\u0026thinsp;+\u0026thinsp;AI: an updated quality, risk of bias, and applicability assessment tool for prediction models using regression or artificial intelligence methods. BMJ. 2025;388:e082505. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1136/bmj-2024-082505\u003c/span\u003e\u003cspan address=\"10.1136/bmj-2024-082505\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParker WH, Feskanich D, Broder MS, Chang E, Shoupe D, Farquhar CM, et al. Long-term mortality associated with oophorectomy compared with ovarian conservation in the nurses\u0026rsquo; health study. Obstet Gynecol. 2013;121:709\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/AOG.0b013e3182864350\u003c/span\u003e\u003cspan address=\"10.1097/AOG.0b013e3182864350\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVannuccini S, Petraglia F, Carmona F, Calaf J, Chapron C. The modern management of uterine fibroids-related abnormal uterine bleeding. Fertil Steril. 2024;122:20\u0026ndash;30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fertnstert.2024.04.041\u003c/span\u003e\u003cspan address=\"10.1016/j.fertnstert.2024.04.041\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePolat G, Arslan HK. Artificial Intelligence in Clinical and Surgical Gynecology. İstanbul Gelişim \u0026Uuml;niversitesi Sağlık. Bilim Derg. 2024;1232\u0026ndash;41. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.38079/igusabder.1291375\u003c/span\u003e\u003cspan address=\"10.38079/igusabder.1291375\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuo T, Li L, Chen X, Wang Z, Zhang X, Liu S, et al. Artificial intelligence-aided method to detect uterine fibroids in ultrasound images: a retrospective study. Sci Rep. 2023;13:3714. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-022-26771-1\u003c/span\u003e\u003cspan address=\"10.1038/s41598-022-26771-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Uterine fibroid, hysterectomy, myomectomy, artificial intelligence, machine learning, anemia parameters, ferritin, hemoglobin, uterine fibroid characteristics, decision support systems","lastPublishedDoi":"10.21203/rs.3.rs-8459924/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8459924/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eUterine fibroids (UFs) are common benign tumors in reproductive-age women, often requiring surgical treatment. The choice between hysterectomy and myomectomy depends on fibroid features and hematologic indices but is usually based on subjective clinical judgment. Advances in artificial intelligence (AI) and machine learning (ML) now enable data-driven decision support systems that can improve surgical accuracy, consistency, and patient outcomes. To develop and validate ML-based models capable of predicting the appropriate surgical approach\u0026mdash;hysterectomy or myomectomy\u0026mdash;using anemia-related laboratory parameters and uterine fibroid characteristics.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eA retrospective multicenter study was conducted across three tertiary referral hospitals. A total of 600 women diagnosed with UFs were included, of whom 362 (60.3%) underwent hysterectomy and 238 (39.7%) underwent myomectomy. Clinical and laboratory data, including fibroid number, total fibroid volume, hemoglobin, and ferritin levels, were analyzed. Comparative statistical analyses were performed, and 126 ML models were trained and tested to predict surgical type based on these variables. A cohort of 50 cases was used for blinded real-time validation, and concordance was assessed.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eFerritin levels, fibroid count, and total fibroid volume were significantly higher in the hysterectomy group compared with the myomectomy group (P\u0026thinsp;\u0026lt;\u0026thinsp;.001 for all). The ML models achieved accuracy rates exceeding 90% in differentiating between the two surgical approaches, effectively replicating clinicians\u0026rsquo; decision-making behavior. Prospective validation demonstrated a high level of agreement, with 96% concordance between ML predictions and the blinded gynecologist\u0026rsquo;s assessments.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eAnemia-related indices and fibroid characteristics emerged as principal factors influencing surgical decision-making in uterine fibroid management. Within an academically designed and prospectively validated framework, the ML\u0026ndash;based model demonstrated high predictive performance and robust concordance with expert clinical assessments, supporting its methodological reliability. These findings indicate that such validated, data-driven algorithms may, following further external validation and implementation studies, be suitable for integration into gynecologic surgical workflows, where they could assist objective preoperative planning and enhance future clinical decision-making.\u003c/p\u003e","manuscriptTitle":"Data-Driven Clinical Decision Support for Predicting Surgical Choice in Uterine Fibroid Management Using Anemia Indices and Fibroid Characteristics","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-08 02:29:28","doi":"10.21203/rs.3.rs-8459924/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ec31deb4-9a9d-4593-a87e-4b5a073ed70a","owner":[],"postedDate":"January 8th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-08T13:47:30+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-08 02:29:28","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8459924","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8459924","identity":"rs-8459924","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}