Predicting the therapeutic efficacy of uterine artery embolization for adenomyosis using a combined model based on MRI radiomics and clinical characteristics

Objective Prediction of the therapeutic efficacy of uterine artery embolization (UAE) for adenomyosis (AM) using an MRI-based radiomics model combined with clinical characteristics.

A retrospective analysis was conducted on 126 patients with AM who underwent UAE at the Interventional Radiology Department of the Second Affiliated Hospital of Soochow University. Radiomics features were extracted from uterine lesions using axial T2-weighted imaging with fat suppression (T2WI-FS) sequences obtained prior to treatment. Following feature selection using the mRMR and LASSO algorithms, radiomics models were developed to predict the lesion necrosis rate in AM after UAE. These models employed the following classifiers: C-SVC, Nu-SVC, Logistic Regression (LR), Random Forest (RF), AdaBoost, and XGBoost. The optimal radiomics model was subsequently identified through receiver operating characteristic (ROC) curve analysis. Relevant clinical characteristics were screened using univariate logistic regression analysis and the LASSO algorithm to identify variables for constructing the clinical model. Finally, a combined model integrating radiomics features and clinical characteristics was developed. The dataset was partitioned into training (n = 100) and test (n = 26) sets at an 8:2 ratio. The predictive performance of the models was evaluated using ROC curves, while their clinical utility was assessed through decision curve analysis (DCA).

Total of 1874 radiomics features were extracted from axial T2WI-FS sequences. Following dimensionality reduction and feature selection, 14 radiomics features were identified as valuable. Among the radiomics models, the RF model demonstrated the highest predictive performance and generalizability, achieving AUC values of 0.796 in the training set and 0.740 in the test set. Subsequently, a clinical model was constructed using clinical characteristics, with the RF model exhibiting superior predictive performance and generalizability, yielding AUC of 0.876 (training set) and 0.817 (test set). Ultimately, the combined model integrating radiomics features and clinical characteristics demonstrated optimal predictive ability. The LR model achieved an AUC of 0.944 in the training set and 0.870 in the test set, while DCA confirmed its optimal clinical utility.

The combined model integrating radiomics features and clinical characteristics demonstrated significant predictive performance and robustness in evaluating lesion necrosis extent following UAE for AM. Its discriminative capability surpassed that of single-modality prediction models, potentially offering a non-invasive objective assessment tool to optimize clinical decision-making pathways. Similar content being viewed by others Data availability The data that support the findings of this study are available from the corresponding author upon reasonable request.

Li Q, Ding Y, Zhang XY, Feng WW, Hua KQ (2018) Drug therapy for adenomyosis: a prospective, nonrandomized, parallel-controlled study. J Int Med Res 46(5):1855–1865. https://doi.org/10.1177/0300060517752997 Koninckx PR, Ussia A, Adamyan L, Tahlak M, Keckstein J, Wattiez A et al (2021) The epidemiology of endometriosis is poorly known as the pathophysiology and diagnosis are unclear. Best Pract Res Clin Obstet Gynaecol 71:14–26. https://doi.org/10.1016/j.bpobgyn.2020.08.005 Ozen M, Patel R, Hoffman M, Raissi D (2023) Update on Endovascular Therapy for Fibroids and Adenomyosis. Semin Intervent Radiol 40(4):327–334. https://doi.org/10.1055/s-0043-1770713 Manyonda I, Belli AM, Lumsden MA, Moss J, Mckinnon W, Middleton LJ et al (2020) Uterine-Artery Embolization or Myomectomy for Uterine Fibroids. N Engl J Med 383(5):440–451. https://doi.org/10.1056/NEJMoa1914735 Balamurugan S, Shah R, Panganiban K, Lehrack M, Agrawal DK (2025) Uterine Artery Embolization: A Growing Pillar of Gynecological Intervention. J Radiol Clin Imaging 8(1):1–17. https://doi.org/10.26502/jrci.2809105 Wentao J, Guofu Z, Tianpin W, Shijia W, Haiyan Z, Wentao L (2022) Nomogram for predicting the long-term outcomes of uterine artery embolization for adenomyosis. Eur J Radiol 148:110183. https://doi.org/10.1016/j.ejrad.2022.110183 Moldassarina RS (2023) Modern view on the diagnostics and treatment of adenomyosis. Arch Gynecol Obstet 308(1):171–181. https://doi.org/10.1007/s00404-023-06982-1 Harmsen MJ, Trommelen LM, de Leeuw RA, Tellum T, Juffermans L, Griffioen AW et al (2023) Uterine junctional zone and adenomyosis: comparison of MRI, transvaginal ultrasound and histology. Ultrasound Obstet Gynecol 62(1):42–60. https://doi.org/10.1002/uog.26117 Meng X, Yang D, Jin H, Xu H, Lu J, Liu Z et al (2025) MRI-based radiomics model for predicting endometrial cancer with high tumor mutation burden. Abdom Radiol (Ny) 50(4):1822–1830. https://doi.org/10.1007/s00261-024-04547-7 Kobayashi H, Matsubara S (2020) A Classification Proposal for Adenomyosis Based on Magnetic Resonance Imaging. Gynecol Obstet Invest 85(2):118–126. https://doi.org/10.1159/000505690 Ananthakrishnan G, Macnaught G, Hinksman L, Gilmour H, Forbes KP, Moss JG (2012) Diffusion-weighted imaging in uterine artery embolisation: do findings correlate with contrast enhancement and volume reduction? Br J Radiol 85(1019):e1046-e1050. https://doi.org/10.1259/bjr/91540172 Kroencke TJ, Scheurig C, Poellinger A, Gronewold M, Hamm B (2010) Uterine artery embolization for leiomyomas: percentage of infarction predicts clinical outcome. Radiology 255(3):834–841. https://doi.org/10.1148/radiol.10090977 Zhou Z, Qian X, Hu J, Chen G, Zhang C, Zhu J et al (2023) An artificial intelligence-assisted diagnosis modeling software (AIMS) platform based on medical images and machine learning: a development and validation study. Quant Imaging Med Surg 13(11):7504–7522. https://doi.org/10.21037/qims-23-20 Kosai S, Higashihara H, Yano H, Kashiwagi E, Nagai K, Tanaka K et al (2023) Risk Factors Associated with Clinical Failure of Uterine Artery Embolization for Postpartum Hemorrhage. J Vasc Interv Radiol 34(1):95–101. https://doi.org/10.1016/j.jvir.2022.09.018 Bohiltea R, Ducu I, Mihai B, Iordache AM, Dorobat B, Vladareanu EM et al (2021) Uterine artery embolization combined with subsequent suction evacuation as low-risk treatment for cesarean scar pregnancy. Diagnostics (Basel). https://doi.org/10.3390/diagnostics11122350 Zeccola AM, Allen SE (2024) Alternative treatments of adenomyosis - an update in procedural management and clinical outcomes. Curr Opin Obstet Gynecol 36(4):287–295. https://doi.org/10.1097/GCO.0000000000000972 Wang S, Meng X, Dong Y (2016) The evaluation of uterine artery embolization as a nonsurgical treatment option for adenomyosis. Int J Gynaecol Obstet 133(2):202–205. https://doi.org/10.1016/j.ijgo.2015.09.016 Singh R (2014) Uterine artery embolization: A soccour for bleeding uterus. Int J Appl Basic Med Res 4(1):1–2. https://doi.org/10.4103/2229-516X.125664 Kim SH, Han K, Shin JH, Kim J, Hyun DH, Lee MY (2023) Repeat uterine artery embolization (UAE) for recurrent postpartum hemorrhage in patients who underwent UAE after a previous delivery: a multicenter study. Eur Radiol 33(7):5037–5044. https://doi.org/10.1007/s00330-023-09440-3 Laschke MW, Elitzsch A, Vollmar B, Vajkoczy P, Menger MD (2006) Combined inhibition of vascular endothelial growth factor (VEGF), fibroblast growth factor and platelet-derived growth factor, but not inhibition of VEGF alone, effectively suppresses angiogenesis and vessel maturation in endometriotic lesions. Hum Reprod 21(1):262–268. https://doi.org/10.1093/humrep/dei308 Kim MD, Kim YM, Kim HC, Cho JH, Kang HG, Lee C et al (2011) Uterine artery embolization for symptomatic adenomyosis: a new technical development of the 1-2-3 protocol and predictive factors of MR imaging affecting outcomes. J Vasc Interv Radiol 22(4):497–502. https://doi.org/10.1016/j.jvir.2011.01.426 Jung DC, Kim MD, Oh YT, Won JY, Lee DY (2012) Prediction of early response to uterine arterial embolisation of adenomyosis: value of T2 signal intensity ratio of adenomyosis. Eur Radiol 22(9):2044–2049. https://doi.org/10.1007/s00330-012-2436-z Lee SH, Hahn ST, Choi BG (2008) Transcatheter arterial embolization for massive hemoptysis in patients with coal workers’ pneumoconiosis: an 11-year experience. Acta Radiol 49(1):26–31. https://doi.org/10.1080/02841850701675701 Kirby JM, Burrows D, Haider E, Maizlin Z, Midia M (2011) Utility of MRI before and after uterine fibroid embolization: why to do it and what to look for. Cardiovasc Intervent Radiol 34(4):705–716. https://doi.org/10.1007/s00270-010-0029-2 Kunachowicz D, Tomecka P, Sedzik M, Kalinin J, Kuznicki J, Rembialkowska N (2025) Influence of hypoxia on tumor heterogeneity, DNA repair, and cancer therapy: from molecular insights to therapeutic strategies. Cells. https://doi.org/10.3390/cells14141057 Charkhchi P, Cybulski C, Gronwald J, Wong FO, Narod SA, Akbari MR (2020) CA125 and ovarian cancer: a comprehensive review. Cancers (Basel). https://doi.org/10.3390/cancers12123730 Kil K, Chung JE, Pak HJ, Jeung IC, Kim JH, Jo HH et al (2015) Usefulness of CA125 in the differential diagnosis of uterine adenomyosis and myoma. Eur J Obstet Gynecol Reprod Biol 185:131–135. https://doi.org/10.1016/j.ejogrb.2014.12.008 Tang Y, Wen MB, Xiang RM, Yang MT, Shu B, Xu F et al (2023) Serum CA125 as a biomarker for dysmenorrhea in adenomyosis. Int J Gynaecol Obstet 163(1):131–139. https://doi.org/10.1002/ijgo.14832 Zheng R, Zeng D, Wan TT, Guo WB, Guo Y, Liu MJ (2018) Predisposing factors for predicting the therapeutic response of adenomyosis after uterine artery embolization: serum CA125 levels and accompanying endometriosis. Diagn Interv Radiol 24(6):364–371. https://doi.org/10.5152/dir.2018.17479 Yu W, Liu G, Liu C, Zhang Z (2018) Recurrence-associated factors of laparoscopic adenomyomectomy for severely symptomatic adenomyoma. Oncol Lett 16(3):3430–3438. https://doi.org/10.3892/ol.2018.9082 Chen H, Liu Y, Zhao J, Jia X, Chai F, Peng Y et al (2024) Quantification of intratumoral heterogeneity using habitat-based MRI radiomics to identify HER2-positive, -low and -zero breast cancers: a multicenter study. Breast Cancer Res 26(1):160. https://doi.org/10.1186/s13058-024-01921-7 Han X, Guo Y, Ye H, Chen Z, Hu Q, Wei X et al (2024) Development of a machine learning-based radiomics signature for estimating breast cancer TME phenotypes and predicting anti-PD-1/PD-L1 immunotherapy response. Breast Cancer Res 26(1):18. https://doi.org/10.1186/s13058-024-01776-y Aerts HJ, Velazquez ER, Leijenaar RT, Parmar C, Grossmann P, Carvalho S et al (2014) Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat Commun 5:4006. https://doi.org/10.1038/ncomms5006 Gillies RJ, Kinahan PE, Hricak H (2016) Radiomics: Images Are More than Pictures, They Are Data. Radiology 278(2):563–577. https://doi.org/10.1148/radiol.2015151169 Author information Authors and Affiliations Contributions HZ, HRC, and ZY contributed equally to this work. HZ and HRC were responsible for data curation, data analysis and manuscript drafting of the manuscript. ZY performed medical image processing and radiomic feature extraction. DW, XKS, and LY contributed to statistical analysis, experimental design, validation of results, and verification of data collection. CG and YJ conceptualized and supervised the project, provided funding acquisition, and approved the final manuscript. XWS participated in the critical review, editing, and intellectual refinement of the manuscript. All authors read and approved the submitted version. Corresponding author Ethics declarations Competing interests The authors declare no competing interests. Additional information Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary Information Below is the link to the electronic supplementary material. Rights and permissions Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law. About this article Cite this article Zhang, H., Ye, Z., Cai, H. et al. Predicting the therapeutic efficacy of uterine artery embolization for adenomyosis using a combined model based on MRI radiomics and clinical characteristics. Abdom Radiol 51, 2093–2105 (2026). https://doi.org/10.1007/s00261-025-05176-4 Received: Revised: Accepted: Published: Version of record: Issue date: DOI: https://doi.org/10.1007/s00261-025-05176-4