Share：
Share this content in WeChat
X
[Chinese][PDF]161
• Clinical Article •
Developing an interpretable model to predict hemorrhagic transformation risk in acute stroke using multiparameter MRI
YU Huihua  JIANG Liang  PENG Mingyang  GENG Wen  Yin Xindao  ZHOU Chunyan 

Cite this article as: YU H H, JIANG L, PENG M Y, et al. Developing an interpretable model to predict hemorrhagic transformation risk in acute stroke using multiparameter MRI[J]. Chin J Magn Reson Imaging, 2025, 16(4): 19-24. DOI:10.12015/issn.1674-8034.2025.04.004.


[Abstract] Objective To develop an interpretable model to predict the risk of hemorrhagic transformation after endovascular treatment in acute stroke, utilizing multiparameter MRI.Materials and Methods A retrospective analysis was conducted on 274 patients who presented with acute stroke at our hospital. The assessment of hemorrhagic transformation in these patients was performed using CT or MRI 24 hours post-treatment. Utilize the PyRadiomics software to extract 1143 features from diffusion-weighted imaging and an additional 1143 features from perfusion-weighted imaging, and develop a radiomics score (Radscore) based on these extracted features. Utilize SHapley Additive exPlanations (SHAP) to identify the most pertinent features for model development. Develop an interpretable prediction model for assessing the risk of bleeding conversion by employing six distinct machine learning classifiers: gradient boosting classifier, random forest (RF), eXtreme gradient boosting (XGB), adaptive boosting, Gaussian naive Bayes, and logistic regression. Assess the predictive performance of these machine learning models using receiver operating characteristic (ROC) curves and decision curve analysis (DCA).Results Following feature screening and dimensionality reduction, 15 features demonstrating a strong correlation with the transformation of acute ischemic stroke bleeding were identified. Five clinical variables with statistical differences (age, time from onset to MRI examination, NIHSS score on admission, history of diabetes, and history of atrial fibrillation) and radscore were incorporated into the machine learning model. Among the models evaluated, the RF model exhibited the highest predictive performance, achieving an area under the curve (AUC) of 0.928. When the critical value is set at 0.844, the model demonstrates an accuracy of 85.5%, a sensitivity of 83.0%, and a specificity of 88.2%. DCA indicates that the RF model provides a substantial net benefit in predicting the risk of hemorrhagic transformation in cases of acute stroke.Conclusions The interpretable RF model, which integrates multiparameter MRI radiomics with clinical features, enhances the accuracy of predicting the risk of hemorrhagic transformation following mechanical thrombectomy in acute ischemic stroke. This model offers valuable guidance for early clinical intervention and treatment.
[Keywords] stroke；hemorrhagic transformation；magnetic resonance imaging；radiomics；machine learning

YU Huihua   JIANG Liang   PENG Mingyang   GENG Wen   Yin Xindao   ZHOU Chunyan*  

Department of Radiology, Nanjing First Hospital, Nanjing Medical University, Nanjing 210006, China

Corresponding author: ZHOU C Y, E-mail: 13951802716@163.com

Conflicts of interest   None.

Received  2024-11-11
Accepted  2025-03-10
DOI: 10.12015/issn.1674-8034.2025.04.004
Cite this article as: YU H H, JIANG L, PENG M Y, et al. Developing an interpretable model to predict hemorrhagic transformation risk in acute stroke using multiparameter MRI[J]. Chin J Magn Reson Imaging, 2025, 16(4): 19-24. DOI:10.12015/issn.1674-8034.2025.04.004.

[1]
CONSOLI A, GORY B. Long-term results of mechanical thrombectomy for large ischaemic stroke[J]. Lancet, 2024, 403(10428): 700-701. DOI: 10.1016/S0140-6736(24)00158-2.
[2]
CHEN X R, YAN L X, HOU L, et al. Trends of Stroke Incidence and Mortality From 2015 to 2019 in China[J]. Chinese Circulation Journal, 2024, 39(5): 470-476. DOI: 10.3969/j.issn.1000-3614.2024.05.007.
[3]
SARRAJ A, HASSAN A E, ABRAHAM M G, et al. Endovascular Thrombectomy for Large Ischemic Stroke Across Ischemic Injury and Penumbra Profiles[J]. JAMA, 2024, 331(9): 750-763. DOI: 10.1001/jama.2024.0572.
[4]
Chinese Stroke Society, Neurological Intervention Branch of Chinese Stroke Society, Interventional Group of Stroke Prevention and Control Professional Committee of Chinese Preventive Medical Association. Chinese Guidelines for Endovascular Treatment of Acute Ichaemic Stroke 2023[J]. Chin J Stroke, 2023, 18(6): 684-711. DOI: 10.3969/j.issn.1673-5765.2023.06.010.
[5]
NILES J, BHASIN G, GANTI L. Large right middle cerebral artery stroke with hemorrhagic transformation[J/OL]. Int J Emerg Med, 2024, 17(1): 138 [2024-11-11]. https://pubmed.ncbi.nlm.nih.gov/39367308/. DOI: 10.1186/s12245-024-00739-6.
[6]
BRACARD S, DUCROCQ X, MAS J L, et al. Mechanical thrombectomy after intravenous alteplase versus alteplase alone after stroke (thrace): A randomised controlled trial[J]. Lancet Neurol, 2016, 15(11): 1138-1147. DOI: 10.1016/S1474-4422(16)30177-6.
[7]
LEIGH R, SENERS P, ROUSSEAU V, et al. Blood-brain barrier profile pretreatment is associated with hemorrhagic transformation after endovascular reperfusion[J]. Ann Clin Transl Neurol, 2024, 11(12): 3292-3299. DOI: 10.1002/acn3.52236.
[8]
CHEN P, SUN X, YANG L, et al. Ischemic core volume is associated with hemorrhagic transformation post endovascular thrombectomy[J/OL]. J Stroke Cerebrovasc Dis, 2024, 33(11): 107889 [2024-11-11]. https://www.strokejournal.org/article/S1052-3057(24)00333-1/fulltext. DOI: 10.1016/j.jstrokecerebrovasdis.2024.107889.
[9]
WANG R, PENG M Y, ZHOU X F, et al. Predictors of intracranial hemorrhage after mechanical thrombectomy in acute ischemic stroke[J]. Chin J Magn Reson Imaging, 2021, 12(1): 9-14. DOI: 10.12015/issn.1674-8034.2021.01.003.
[10]
GARCÍA JURADO P B, ROLDÁN ROMERO E, PÉREZ MONTILLA M E, et al. Incidence, prognosis, and prediction of haemorrhagic transformation after revascularisation for stroke[J]. Neurologia (Engl Ed), 2021, 36(8): 589-596. DOI: 10.1016/j.nrleng.2020.05.001.
[11]
MA Y, XU D Y, LIU Q, et al. Nomogram prediction model for the risk of intracranial hemorrhagic transformation after intravenous thrombolysis in patients with acute ischemic stroke[J/OL]. Front Neurol, 2024, 15: 1361035 [2024-11-11]. http://pmc-ncbi-nlm-nih-gov-s.webvpn.njmu.edu.cn:8118/articles/PMC10956578. DOI: 10.3389/fneur.2024.1361035.
[12]
WU X M, KE X J, SUN B, et al. Predictive value of stress hyperglycemia ratio combined with glycosylated hemoglobin on hemor-rhagic transformation after intravenous thrombolysis in elderly patients with acute ischemic stroke[J]. Practical Geriatrics, 2024, 38(1): 54-58. DOI: 10.3969/j.issn.1003-9198.2024.01.013.
[13]
HONG L, HSU T M, ZHANG Y, et al. Neuroimaging Prediction of Hemorrhagic Transformation for Acute Ischemic Stroke[J]. Cerebrovasc Dis, 2022, 51(4): 542-552. DOI: 10.1159/000521150.
[14]
YANG M, TANG L, HU Z, et al. Application of Neuroimaging for the Prediction of Hemorrhagic Transformation after Intravenous Thrombolysis in Acute Ischemic Stroke[J]. Cerebrovasc Dis, 2023, 52(1): 1-10. DOI: 10.1159/000524749.
[15]
NAEL K, KNITTER J R, JAHAN R, et al. Multiparametric Magnetic Resonance Imaging for Prediction of Parenchymal Hemorrhage in Acute Ischemic Stroke After Reperfusion Therapy[J]. Stroke, 2017, 48(3): 664-670. DOI: 10.1161/STROKEAHA.116.014343.
[16]
SUH C H, JUNG S C, CHO S J, et al. MRI for prediction of hemorrhagic transformation in acute ischemic stroke: a systematic review and meta-analysis[J]. Acta Radiol, 2020, 61(7): 964-972. DOI: 10.1177/0284185119887593.
[17]
ANDE S R, GRYNSPAN J, AVIV R I, et al. Imaging for Predicting Hemorrhagic Transformation of Acute Ischemic Stroke-A Narrative Review[J]. Can Assoc Radiol J, 2022, 73(1): 194-202. DOI: 10.1177/08465371211018369.
[18]
GUO X, LIU R, JIA M, et al. Ischemia Reperfusion Injury Induced Blood Brain Barrier Dysfunction and the Involved Molecular Mechanism[J]. Neurochem Res, 2023, 48(8): 2320-2334. DOI: 10.1007/s11064-023-03923-x.
[19]
CHARBONNIER G, BONNET L, BIONDI A, et al. Intracranial Bleeding After Reperfusion Therapy in Acute Ischemic Stroke[J/OL]. Front Neurol, 2021, 11: 629920 [2024-11-11]. http://pmc-ncbi-nlm-nih-gov-s.webvpn.njmu.edu.cn:8118/articles/PMC7900408. DOI: 10.3389/fneur.2020.629920.
[20]
NAGARAJA N. Diffusion weighted imaging in acute ischemic stroke: A review of its interpretation pitfalls and advanced diffusion imaging application[J/OL]. J Neurol Sci, 2021, 425: 117435 [2024-11-11]. https://www.jns-journal.com/article/S0022-510X(21)00129-5/fulltext. DOI: 10.1016/j.jns.2021.117435.
[21]
NEUBERGER U, KICKINGEREDER P, SCHONENBERGER S, et al. Risk factors of intracranial hemorrhage after mechanical thrombectomy of anterior circulation ischemic stroke[J]. Neuroradiology, 2019, 61(4): 461-469. DOI: 10.1001/jama.2024.0572.
[22]
UPADHYAYA P, MEHTA A, LUBY M, et al. Good outcome associated with blood-brain barrier disruption and lower blood pressure after endovascular therapy[J/OL]. J Stroke Cerebrovasc Dis, 2023, 32(1): 106870 [2024-11-11]. http://pmc-ncbi-nlm-nih-gov-s.webvpn.njmu.edu.cn:8118/articles/PMC9825643. DOI: 10.1016/j.jstrokecerebrovasdis.2022.106870.
[23]
CHANG H, LI Y L, WANG L. Clinical Research Progress on the Mechanism of Hemorrhagic Transformation after Intravenous Thrombolysis in Ischemic Stroke[J]. Mod Diagn Treat, 2023, 34(23): 3505-3507, 3515.
[24]
LUO L, HU Q, YAN R, et al. Alpha-Asarone Ameliorates Neuronal Injury After Ischemic Stroke and Hemorrhagic Transformation by Attenuating Blood-Brain Barrier Destruction, Promoting Neurogenesis, and Inhibiting Neuroinflammation[J/OL]. Mol Neurobiol, 2024 [2024-11-11]. http://link-springer-com-s.webvpn.njmu.edu.cn:8118/article/10.1007/s12035-024-04596-5. DOI: 10.1007/s12035-024-04596-5.
[25]
OKAZAKI S, YAMAGAMI H, YOSHIMOTO T, et al. Cerebral hyperperfusion on arterial spin labeling MRI after reperfusion therapy is related to hemorrhagic transformation[J]. J Cereb Blood Flow Metab, 2017, 37(9): 3087-3090. DOI: 10.1177/0271678X17718099.
[26]
RHODES J S, CUTLER A, MOON K R. Geometry- and Accuracy-Preserving Random Forest Proximities[J]. IEEE Trans Pattern Anal Mach Intell, 2023, 45(9): 10947-10959. DOI: 10.1109/TPAMI.2023.3263774.
[27]
GHOSH D, CABRERA J. Enriched Random Forest for High Dimensional Genomic Data[J]. IEEE/ACM Trans Comput Biol Bioinform, 2022, 19(5): 2817-2828. DOI: 10.1109/TCBB.2021.3089417.
[28]
YU Y, GUO D, LOU M, et al. Prediction of hemorrhagic transformation severity in acute stroke from source perfusion MRI[J]. IEEE Trans Biomed Eng, 2018, 65(9): 2058-2065. DOI: 10.1109/TBME.2017.2783241.
[29]
HORIE N, MOROFUJI Y, IKI Y, et al. Impact of basal ganglia damage after successful endovascular recanalization for acute ischemic stroke involving lenticulostriate arteries[J]. J Neurosurg, 2019, 132(6): 1880-1888. DOI: 10.3171/2019.3.JNS182909.
[30]
JIANG L, MIAO Z, CHEN H, et al. Radiomics Analysis of Diffusion-Weighted Imaging and Long-Term Unfavorable Outcomes Risk for Acute Stroke[J]. Stroke, 2023, 54(2): 488-498. DOI: 10.1161/STROKEAHA.122.040418.
[31]
ROUSHDY T, ABDEL NASSER A, NASEF A, et al. A clinical comparative analysis between expanded NIHSS and original NIHSS in posterior circulation ischemic stroke[J]. J Clin Neurosci, 2023, 114: 77-80. DOI: 10.1016/j.jocn.2023.06.006.
[32]
HSIEH C Y, LEE C H, WU D P, et al. Characteristics and outcomes of ischemic stroke in patients with known atrial fibrillation or atrial fibrillation diagnosed after stroke[J]. Int J Cardiol, 2018, 261: 68-72. DOI: 10.1016/j.ijcard.2017.11.047.
[33]
LUNDBERG S M, LEE S I. A unified approach to interpreting model predictions[C]//Advances in Neural Information Processing Systems. New York: Curran Associates, 2017: 4765-4774. https://doi.org/10.48550/arXiv.1705.07874.

PREV Analysis of characteristics of brain functional network in Crohn,s disease patients with chronic abdominal pain based on rs-fMRI graph theory
NEXT Correlation study of atherosclerotic plaques in the middle cerebral artery with recurrent stroke assessed by magnetic resonance vascular wall imaging
  


LINK


Tel & Fax: +8610-67113815    E-mail: editor@cjmri.cn