journal article
Open Access
Mar 29, 2023
A clinical–radiomics model based on noncontrast computed tomography to predict hemorrhagic transformation after stroke by machine learning: a multicenter study
Abstract
Abstract
Objective
To build a clinical–radiomics model based on noncontrast computed tomography images to identify the risk of hemorrhagic transformation (HT) in patients with acute ischemic stroke (AIS) following intravenous thrombolysis (IVT).
Materials and methods
A total of 517 consecutive patients with AIS were screened for inclusion. Datasets from six hospitals were randomly divided into a training cohort and an internal cohort with an 8:2 ratio. The dataset of the seventh hospital was used for an independent external verification. The best dimensionality reduction method to choose features and the best machine learning (ML) algorithm to develop a model were selected. Then, the clinical, radiomics and clinical–radiomics models were developed. Finally, the performance of the models was measured using the area under the receiver operating characteristic curve (AUC).
Results
Of 517 from seven hospitals, 249 (48%) had HT. The best method for choosing features was recursive feature elimination, and the best ML algorithm to build models was extreme gradient boosting. In distinguishing patients with HT, the AUC of the clinical model was 0.898 (95% CI 0.873–0.921) in the internal validation cohort, and 0.911 (95% CI 0.891–0.928) in the external validation cohort; the AUC of radiomics model was 0.922 (95% CI 0.896–0.941) and 0.883 (95% CI 0.851–0.902), while the AUC of clinical–radiomics model was 0.950 (95% CI 0.925–0.967) and 0.942 (95% CI 0.927–0.958) respectively.
Conclusion
The proposed clinical–radiomics model is a dependable approach that could provide risk assessment of HT for patients who receive IVT after stroke.
Objective
To build a clinical–radiomics model based on noncontrast computed tomography images to identify the risk of hemorrhagic transformation (HT) in patients with acute ischemic stroke (AIS) following intravenous thrombolysis (IVT).
Materials and methods
A total of 517 consecutive patients with AIS were screened for inclusion. Datasets from six hospitals were randomly divided into a training cohort and an internal cohort with an 8:2 ratio. The dataset of the seventh hospital was used for an independent external verification. The best dimensionality reduction method to choose features and the best machine learning (ML) algorithm to develop a model were selected. Then, the clinical, radiomics and clinical–radiomics models were developed. Finally, the performance of the models was measured using the area under the receiver operating characteristic curve (AUC).
Results
Of 517 from seven hospitals, 249 (48%) had HT. The best method for choosing features was recursive feature elimination, and the best ML algorithm to build models was extreme gradient boosting. In distinguishing patients with HT, the AUC of the clinical model was 0.898 (95% CI 0.873–0.921) in the internal validation cohort, and 0.911 (95% CI 0.891–0.928) in the external validation cohort; the AUC of radiomics model was 0.922 (95% CI 0.896–0.941) and 0.883 (95% CI 0.851–0.902), while the AUC of clinical–radiomics model was 0.950 (95% CI 0.925–0.967) and 0.942 (95% CI 0.927–0.958) respectively.
Conclusion
The proposed clinical–radiomics model is a dependable approach that could provide risk assessment of HT for patients who receive IVT after stroke.
Topics
No keywords indexed for this article. Browse by subject →
References
42
[1]
Pandian JD, Gall SL, Kate MP et al (2018) Prevention of stroke: a global perspective. Lancet 392:1269–1278. https://doi.org/10.1016/S0140-6736(18)31269-8
10.1016/s0140-6736(18)31269-8
[2]
Powers WJ, Rabinstein AA, Ackerson T et al (2019) Guidelines for the early management of patients with acute ischemic stroke: 2019 update to the 2018 guidelines for the early management of acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 50:e344–e418. https://doi.org/10.1161/STR.0000000000000211
10.1161/str.0000000000000211
[3]
Xing Y, Guo ZN, Yan S et al (2014) Increased globulin and its association with hemorrhagic transformation in patients receiving intra-arterial thrombolysis therapy. Neurosci Bull 30:469–476. https://doi.org/10.1007/s12264-013-1440-x
10.1007/s12264-013-1440-x
[4]
Effect of treatment delay, age, and stroke severity on the effects of intravenous thrombolysis with alteplase for acute ischaemic stroke: a meta-analysis of individual patient data from randomised trials
Jonathan Emberson, Kennedy R Lees, Patrick Lyden et al.
The Lancet
2014
10.1016/s0140-6736(14)60584-5
[5]
Wang Y, Cheng Y, Song Q et al (2020) The association between monocyte to high-density lipoprotein ratio and hemorrhagic transformation in patients with acute ischemic stroke. Aging 12:2498–2506. https://doi.org/10.18632/aging.102757
10.18632/aging.102757
[6]
Liu C, Zhang Y, Niu L, Li J (2022) High level of the fibrin degradation products at admission predicts parenchymal hematoma and unfavorable outcome of ischemic stroke after intravenous thrombolysis. Front Neurol 12:797394. https://doi.org/10.3389/fneur.2021.797394
10.3389/fneur.2021.797394
[7]
Saposnik G, Demchuk A, Tu JV et al (2013) The iScore predicts efficacy and risk of bleeding in the National Institute of Neurological disorders and Stroke Tissue Plasminogen Activator Stroke Trial. J Stroke Cerebrovasc Dis 22:876–882. https://doi.org/10.1016/j.jstrokecerebrovasdis.2012.09.001
10.1016/j.jstrokecerebrovasdis.2012.09.001
[8]
Kim T, Koo J, Kim SH et al (2018) Blood-brain barrier permeability assessed by perfusion computed tomography predicts hemorrhagic transformation in acute reperfusion therapy. Neurol Sci 39:1579–1584. https://doi.org/10.1007/s10072-018-3468-1
10.1007/s10072-018-3468-1
[9]
El Nawar R, Yeung J, Labreuche J et al (2019) MRI-based predictors of hemorrhagic transformation in patients with stroke treated by intravenous thrombolysis. Front Neurol 10:897. https://doi.org/10.3389/fneur.2019.00897
10.3389/fneur.2019.00897
[10]
Jain AR, Jain M, Kanthala AR et al (2013) Association of CT perfusion parameters with hemorrhagic transformation in acute ischemic stroke. AJNR Am J Neuroradiol 34:1895–1900. https://doi.org/10.3174/ajnr.A3502
10.3174/ajnr.a3502
[11]
Yassi N, Parsons MW, Christensen S et al (2013) Prediction of poststroke hemorrhagic transformation using computed tomography perfusion. Stroke 44:3039–3043. https://doi.org/10.1161/STROKEAHA.113.002396
10.1161/strokeaha.113.002396
[12]
Campbell BC, Christensen S, Butcher KS et al (2010) Regional very low cerebral blood volume predicts hemorrhagic transformation better than diffusion-weighted imaging volume and thresholded apparent diffusion coefficient in acute ischemic stroke. Stroke 41:82–88. https://doi.org/10.1161/STROKEAHA.109.562116
10.1161/strokeaha.109.562116
[13]
Oppenheim C, Samson Y, Dormont D et al (2002) DWI prediction of symptomatic hemorrhagic transformation in acute MCA infarct. J Neuroradiol 29:6–13
[14]
Shinoda N, Hori S, Mikami K et al (2017) Prediction of hemorrhagic transformation after acute thrombolysis following major artery occlusion using relative ADC ratio: a retrospective study. J Neuroradiol 44:361–366. https://doi.org/10.1016/j.neurad.2017.07.003
10.1016/j.neurad.2017.07.003
[15]
Mazya MV, Bovi P, Castillo J et al (2013) External validation of the SEDAN score for prediction of intracerebral hemorrhage in stroke thrombolysis. Stroke 44:1595–1600. https://doi.org/10.1161/STROKEAHA.113.000794
10.1161/strokeaha.113.000794
[16]
Lambin P, Rios-Velazquez E, Leijenaar R et al (2012) Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer 48:441–446. https://doi.org/10.1016/j.ejca.2011.11.036
10.1016/j.ejca.2011.11.036
[17]
Kumar V, Gu Y, Basu S et al (2012) Radiomics: the process and the challenges. Magn Reson Imaging 30:1234–1248. https://doi.org/10.1016/j.mri.2012.06.010
10.1016/j.mri.2012.06.010
[18]
Chen Q, Zhu D, Liu J et al (2021) Clinical-radiomics nomogram for risk estimation of early hematoma expansion after acute intracerebral hemorrhage. Acad Radiol 28:307–317. https://doi.org/10.1016/j.acra.2020.02.021
10.1016/j.acra.2020.02.021
[19]
Xu W, Ding Z, Shan Y et al (2020) A nomogram model of radiomics and satellite sign number as imaging predictor for intracranial hematoma expansion. Front Neurosci 14:491. https://doi.org/10.3389/fnins.2020.00491
10.3389/fnins.2020.00491
[20]
Lin J, Li X, Wu G et al (2020) White matter high signals interfere with noncontrast computed tomography in the early identification of cerebral infarction. Cerebrovasc Dis 49:135–143. https://doi.org/10.1159/000505807
10.1159/000505807
[21]
Wen X, Shu Z, Li Y et al (2021) Developing a model for estimating infarction onset time based on computed tomography radiomics in patients with acute middle cerebral artery occlusion. BMC Med Imaging 21:147. https://doi.org/10.1186/s12880-021-00678-1
10.1186/s12880-021-00678-1
[22]
Yu Y, Guo D, Lou M et al (2018) Prediction of hemorrhagic transformation severity in acute stroke from source perfusion MRI. IEEE Trans Biomed Eng 65:2058–2065. https://doi.org/10.1109/TBME.2017.2783241
10.1109/tbme.2017.2783241
[23]
Lee H, Lee EJ, Ham S et al (2020) Machine learning approach to identify stroke within 4.5 hours. Stroke 51:860–866. https://doi.org/10.1161/STROKEAHA.119.027611
10.1161/strokeaha.119.027611
[24]
Wang F, Huang Y, Xia Y et al (2020) Personalized risk prediction of symptomatic intracerebral hemorrhage after stroke thrombolysis using a machine-learning model. Ther Adv Neurol Disord 13:1756286420902358. https://doi.org/10.1177/1756286420902358
10.1177/1756286420902358
[25]
Bentley P, Ganesalingam J, Carlton Jones AL et al (2014) Prediction of stroke thrombolysis outcome using CT brain machine learning. NeuroImage Clin 4:635–640. https://doi.org/10.1016/j.nicl.2014.02.003
10.1016/j.nicl.2014.02.003
[26]
Hacke W, Kaste M, Fieschi C et al (1998) Randomised double-blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS II). Second European-Australasian Acute Stroke Study Investigators. Lancet 352:1245–1251. https://doi.org/10.1016/s0140-6736(98)08020-9
10.1016/s0140-6736(98)08020-9
[27]
Xu X, Li C, Wan T et al (2017) Risk factors for hemorrhagic transformation after intravenous thrombolysis in acute cerebral infarction: a retrospective single-center study. World Neurosurg 101:155–160. https://doi.org/10.1016/j.wneu.2017.01.091
10.1016/j.wneu.2017.01.091
[28]
Kidwell CS, Saver JL, Carneado J et al (2002) Predictors of hemorrhagic transformation in patients receiving intra-arterial thrombolysis. Stroke 33:717–724. https://doi.org/10.1161/hs0302.104110
10.1161/hs0302.104110
[29]
Liu M, Pan Y, Zhou L et al (2017) Predictors of post-thrombolysis symptomatic intracranial hemorrhage in Chinese patients with acute ischemic stroke. PLoS One 12:e0184646. https://doi.org/10.1371/journal.pone.0184646
10.1371/journal.pone.0184646
[30]
Trouillas P, Derex L, Philippeau F et al (2004) Early fibrinogen degradation coagulopathy is predictive of parenchymal hematomas in cerebral rt-PA thrombolysis: a study of 157 cases. Stroke 35:1323–1328. https://doi.org/10.1161/01.STR.0000126040.99024.cf
10.1161/01.str.0000126040.99024.cf
[31]
Sun X, Berthiller J, Trouillas P et al (2015) Early fibrinogen degradation coagulopathy: a predictive factor of parenchymal hematomas in cerebral rt-PA thrombolysis. J Neurol Sci 351:109–114. https://doi.org/10.1016/j.jns.2015.02.048
10.1016/j.jns.2015.02.048
[32]
Verni CC, Davila A, Sims CA et al (2020) D-dimer and fibrin degradation products impair platelet signaling: plasma D-dimer is a predictor and mediator of platelet dysfunction during trauma. J Appl Lab Med 5:1253–1264. https://doi.org/10.1093/jalm/jfaa047
10.1093/jalm/jfaa047
[33]
Jickling GC, Liu D, Stamova B et al (2014) Hemorrhagic transformation after ischemic stroke in animals and humans. J Cereb Blood Flow Metab 34:185–199. https://doi.org/10.1038/jcbfm.2013.203
10.1038/jcbfm.2013.203
[34]
Liberale L, Montecucco F, Bonaventura A et al (2017) Monocyte count at onset predicts poststroke outcomes during a 90-day follow-up. Eur J Clin Invest 47:702–710. https://doi.org/10.1111/eci.12795
10.1111/eci.12795
[35]
Daniels LB, Maisel AS (2007) Natriuretic peptides. J Am Coll Cardiol 50:2357–2368. https://doi.org/10.1016/j.jacc.2007.09.021
10.1016/j.jacc.2007.09.021
[36]
Saposnik G, Gladstone D, Raptis R et al (2013) Atrial fibrillation in ischemic stroke: predicting response to thrombolysis and clinical outcomes. Stroke 44:99–104. https://doi.org/10.1161/STROKEAHA.112.676551
10.1161/strokeaha.112.676551
[37]
Vaclavik D, Vilionskis A, Jatuzis D et al (2018) Clinical outcome of cardioembolic stroke treated by intravenous thrombolysis. Acta Neurol Scand 137:347–355. https://doi.org/10.1111/ane.12880
10.1111/ane.12880
[38]
Zhang KJ, Jin H, Xu R et al (2021) N-terminal pro-brain natriuretic peptide is associated with hemorrhagic transformation and poor outcomes in patients with stroke treated with intravenous thrombolysis. Front Mol Neurosci 14:758915. https://doi.org/10.3389/fnmol.2021.758915
10.3389/fnmol.2021.758915
[39]
Hajdinjak E, Klemen P, Grmec S (2012) Prognostic value of a single prehospital measurement of N-terminal pro-brain natriuretic peptide and troponin T after acute ischaemic stroke. J Int Med Res 40:768–776. https://doi.org/10.1177/147323001204000243
10.1177/147323001204000243
[40]
Paciaroni M, Agnelli G, Corea F et al (2008) Early hemorrhagic transformation of brain infarction: rate, predictive factors, and influence on clinical outcome: results of a prospective multicenter study. Stroke 39:2249–2256. https://doi.org/10.1161/STROKEAHA.107.510321
10.1161/strokeaha.107.510321
[41]
Kalinin MN, Khasanova DR, Ibatullin MM (2017) The hemorrhagic transformation index score: a prediction tool in middle cerebral artery ischemic stroke. BMC Neurol 17:177. https://doi.org/10.1186/s12883-017-0958-3
10.1186/s12883-017-0958-3
[42]
Meyer M, Ronald J, Vernuccio F et al (2019) Reproducibility of CT radiomic features within the same patient: influence of radiation dose and CT reconstruction settings. Radiology 293(3):583–591. https://doi.org/10.1148/radiol.2019190928
10.1148/radiol.2019190928
Cited By
32
Artificial intelligence in ischemic stroke images: current applications and future directions
Frontiers in Neurology
Ying Liu, Zhongjian Wen · 2024
Metrics
32
Citations
42
References
Details
- Published
- Mar 29, 2023
- Vol/Issue
- 14(1)
- License
- View
Authors
Funding
Chunhui Project Foundation of the Education Department of China
Award: Y2020MSXM07
Medical Research Program of the Chongqing National Health Commission and Chongqing Science and Technology Bureau
Award: 2021MSXM155
Heino och Sigrid Jänes Stiftelse för Vetenskaplig Grund- och Forskarutbildning i Geovetenskap
Award: CYS22555
Cite This Article
Huanhuan Ren, Haojie Song, Jingjie Wang, et al. (2023). A clinical–radiomics model based on noncontrast computed tomography to predict hemorrhagic transformation after stroke by machine learning: a multicenter study. Insights into Imaging, 14(1). https://doi.org/10.1186/s13244-023-01399-5
Related
You May Also Like
Convolutional neural networks: an overview and application in radiology
Rikiya Yamashita, Mizuho Nishio · 2018
3,764 citations
Radiomics in medical imaging—“how-to” guide and critical reflection
Janita E. van Timmeren, Davide Cester · 2020
1,092 citations
Artificial intelligence as a medical device in radiology: ethical and regulatory issues in Europe and the United States
Filippo Pesapane, Caterina Volonté · 2018
394 citations
Attitudes and perceptions of UK medical students towards artificial intelligence and radiology: a multicentre survey
Cherry Sit, Rohit Srinivasan · 2020
392 citations
Detector-based spectral CT with a novel dual-layer technology: principles and applications
Negin Rassouli, Maryam Etesami · 2017
241 citations