Determining acute ischemic stroke onset time using machine learning and radiomics features of infarct lesions and whole brain

Jiaxi Lu; Yingwei Guo; Mingming Wang; Yu Luo; Xueqiang Zeng; Xiaoqiang Miao; Asim Zaman; Huihui Yang; Anbo Cao; Yan Kang; Jiaxi Lu; Yingwei Guo; Mingming Wang; Yu Luo; Xueqiang Zeng; Xiaoqiang Miao; Asim Zaman; Huihui Yang; Anbo Cao; Yan Kang

doi:10.3934/mbe.2024002

Mathematical Biosciences and Engineering

2024, Volume 21, Issue 1: 34-48. doi: 10.3934/mbe.2024002

Previous Article Next Article

Research article Special Issues

Determining acute ischemic stroke onset time using machine learning and radiomics features of infarct lesions and whole brain

1.
School of Applied Technology, Shenzhen University, Shenzhen 518060, China
2.
College of Health Science and Environmental Engineering, Shenzhen Technology University, Shenzhen 518118, China
3.
College of Medicine and Biological Information Engineering, Northeastern University, Shenyang 110169, China
4.
Department of Radiology, Shanghai Fourth People's Hospital Affiliated to Tongji University School of Medicine, Shanghai 200434, China
5.
Engineering Research Centre of Medical Imaging and Intelligent Analysis, Ministry of Education, Shenyang 110169, China

Academic Editor: Diego Oliva

Received: 15 August 2023 Revised: 23 November 2023 Accepted: 26 November 2023 Published: 08 December 2023

Accurate determination of the onset time in acute ischemic stroke (AIS) patients helps to formulate more beneficial treatment plans and plays a vital role in the recovery of patients. Considering that the whole brain may contain some critical information, we combined the Radiomics features of infarct lesions and whole brain to improve the prediction accuracy. First, the radiomics features of infarct lesions and whole brain were separately calculated using apparent diffusion coefficient (ADC), diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) sequences of AIS patients with clear onset time. Then, the least absolute shrinkage and selection operator (Lasso) was used to select features. Four experimental groups were generated according to combination strategies: Features in infarct lesions (IL), features in whole brain (WB), direct combination of them (IW) and Lasso selection again after direct combination (IWS), which were used to evaluate the predictive performance. The results of ten-fold cross-validation showed that IWS achieved the best AUC of 0.904, which improved by 13.5% compared with IL (0.769), by 18.7% compared with WB (0.717) and 4.2% compared with IW (0.862). In conclusion, combining infarct lesions and whole brain features from multiple sequences can further improve the accuracy of AIS onset time.
- acute ischemic stroke,
- onset time,
- whole brain,
- machine learning,
- radiomics
Citation: Jiaxi Lu, Yingwei Guo, Mingming Wang, Yu Luo, Xueqiang Zeng, Xiaoqiang Miao, Asim Zaman, Huihui Yang, Anbo Cao, Yan Kang. Determining acute ischemic stroke onset time using machine learning and radiomics features of infarct lesions and whole brain[J]. Mathematical Biosciences and Engineering, 2024, 21(1): 34-48. doi: 10.3934/mbe.2024002

Related Papers:

Abstract

Accurate determination of the onset time in acute ischemic stroke (AIS) patients helps to formulate more beneficial treatment plans and plays a vital role in the recovery of patients. Considering that the whole brain may contain some critical information, we combined the Radiomics features of infarct lesions and whole brain to improve the prediction accuracy. First, the radiomics features of infarct lesions and whole brain were separately calculated using apparent diffusion coefficient (ADC), diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) sequences of AIS patients with clear onset time. Then, the least absolute shrinkage and selection operator (Lasso) was used to select features. Four experimental groups were generated according to combination strategies: Features in infarct lesions (IL), features in whole brain (WB), direct combination of them (IW) and Lasso selection again after direct combination (IWS), which were used to evaluate the predictive performance. The results of ten-fold cross-validation showed that IWS achieved the best AUC of 0.904, which improved by 13.5% compared with IL (0.769), by 18.7% compared with WB (0.717) and 4.2% compared with IW (0.862). In conclusion, combining infarct lesions and whole brain features from multiple sequences can further improve the accuracy of AIS onset time.

References

[1]	Q. Ding, S. Liu, Y. Yao, H. Liu, T. Cai, L. Han, Global, regional, and national burden of ischemic stroke, 1990–2019, Neurology, 98 (2022), E279–E290. https://doi.org/10.1212/WNL.0000000000013115 doi: 10.1212/WNL.0000000000013115
[2]	S. Ogoh, T. Tarumi, Cerebral blood flow regulation and cognitive function: A role of arterial baroreflex function, J. Physiol. Sci., 69 (2019), 813–823. https://doi.org/10.1007/s12576-019-00704-6 doi: 10.1007/s12576-019-00704-6
[3]	W. Hacke, M. Kaste, E. Bluhmki, M. Brozman, A. Dávalos, D. Guidetti, et al., Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke, New Engl. J. Med., 359 (2008), 1317–1329. https://doi.org/10.1056/NEJMoa0804656 doi: 10.1056/NEJMoa0804656
[4]	H. Liu, W. Hu, F. Zhang, W. Gu, J. Hong, J. Chen, et al., Efficacy and safety of rt-PA intravenous thrombolysis in patients with wake-up stroke: A meta-analysis, Medicine, 101 (2022), e28914. https://doi.org/10.1097%2FMD.0000000000028914
[5]	A. R. Al-Buhairi, M. M. Jan, Recombinant tissue plasminogen activator for acute ischemic stroke, Neurosci. J., 7 (2002), 7–13.
[6]	A. Nelson, G. Kelly, R. Byyny, C. Dionne, C. Preslaski, K. Kaucher, Tenecteplase utility in acute ischemic stroke patients: A clinical review of current evidence, Am. J. Emerg. Med., 37 (2019): 344–348. https://doi.org/10.1016/j.ajem.2018.11.018 doi: 10.1016/j.ajem.2018.11.018
[7]	B. C. Campbell, H. Ma, S. Curtze, G. A. Donnan, M. Kaste, Extending thrombolysis to 4.5–9 h and wake-up stroke using perfusion imaging: a systematic review and meta-analysis of individual patient data, Lancet, 394 (2019), 139–147. https://doi.org/10.1016/S0140-6736(19)31053-0 doi: 10.1016/S0140-6736(19)31053-0
[8]	A. Damiza-Detmer, I. Damiza, M. Pawełczyk, Wake-up stroke-diagnosis, management and treatment, Curr. Neurol., 20 (2020), 66–70. https://doi.org/10.15557/AN.2020.0009 doi: 10.15557/AN.2020.0009
[9]	A. Wouters, R. Lemmens, P. Dupont, V. Thijs, Wake-up stroke and stroke of unknown onset: a critical review, Front. Neurol., 5 (2014). https://doi.org/10.3389/fneur.2014.00153 doi: 10.3389/fneur.2014.00153
[10]	C. S. Anderson, T. Robinson, R. I. Lindley, H. Arima, P. M. Lavados, T. H. Lee, et al., Low-dose versus standard-dose intravenous alteplase in acute ischemic stroke, New Engl. J. Med., 374 (2016), 2313–2323. https://doi.org/10.1056/NEJMoa1515510 doi: 10.1056/NEJMoa1515510
[11]	J. Mackey, D. Kleindorfer, H. Sucharew, C. J. Moomaw, B. M. Kissela, K. Alwell, et al., Population-based study of wake-up strokes, Neurology, 76 (2011), 1662–1667. https://doi.org/10.1212/WNL.0b013e318219fb30 doi: 10.1212/WNL.0b013e318219fb30
[12]	S. Emeriau, I. Serre, O. Toubas, F. Pombourcq, C. Oppenheim, L. Pierot, Can diffusion-weighted imaging-fluid-attenuated inversion recovery mismatch (positive diffusion-weighted imaging/negative fluid-attenuated inversion recovery) at 3 tesla identify patients with stroke at < 4.5 hours?, Stroke, 44 (2013), 1647–1651. https://doi.org/10.1161/STROKEAHA.113.001001 doi: 10.1161/STROKEAHA.113.001001
[13]	D. Buck, L. C. Shaw, C. I. Price, G. A. Ford, Reperfusion therapies for wake-up stroke: systematic review, Stroke, 45 (2014), 1869–1875. https://doi.org/10.1161/STROKEAHA.114.005126 doi: 10.1161/STROKEAHA.114.005126
[14]	O. M. Rø nning, Reperfusion therapy in stroke cases with unknown onset, Tidsskrift for Den norske legeforening, 136 (2016), 1333. https://doi.org/10.4045/tidsskr.16.0626 doi: 10.4045/tidsskr.16.0626
[15]	Q. Chen, T. Xia, M. Zhang, N. Xia, J. Liu, Y. Yang, Radiomics in stroke neuroimaging: techniques, applications, and challenges, Aging Dis., 12 (2021), 143–154. https://doi.org/10.14336%2FAD.2020.0421
[16]	K. C. Ho, W. Speier, H. Zhang, F. Scalzo, S. El-Saden, C. W. Arnold, A machine learning approach for classifying ischemic stroke onset time from imaging, IEEE Trans. Med. Imaging, 38 (2019), 1666–1676. https://doi.org/10.1109/TMI.2019.2901445 doi: 10.1109/TMI.2019.2901445
[17]	H. Lee, E. J. Lee, S. Ham, H. B. Lee, J. S. Lee, S. U. Kwon, et al., Machine learning approach to identify stroke within 4.5 hours, Stroke, 51 (2020), 860–866. https://doi.org/10.1161/STROKEAHA.119.027611 doi: 10.1161/STROKEAHA.119.027611
[18]	H. Zhu, L. Jiang, H. Zhang, L. Luo, Y. Chen, Y. Chen, An automatic machine learning approach for ischemic stroke onset time identification based on DWI and FLAIR imaging, Neuroimage-Clin., 31 (2021), 102744. https://doi.org/10.1016/j.nicl.2021.102744 doi: 10.1016/j.nicl.2021.102744
[19]	Y. Q. Zhang, A. F. Liu, F. Y. Man, Y. Y. Zhang, C. Li, Y. E. Liu, et al., MRI radiomic features-based machine learning approach to classify ischemic stroke onset time, J. Neurol., 269 (2022), 350–360.
[20]	M. Jenkinson, S. Smith, A global optimisation method for robust affine registration of brain images, Med. Image Anal., 5 (2001), 143–156. https://doi.org/10.1016/S1361-8415(01)00036-6 doi: 10.1016/S1361-8415(01)00036-6
[21]	M. M. Jenkinson, P. Bannister, M. Brady, S. Smith, Improved optimisation for the robust and accurate linear registration and motion correction of brain images, Neuroimage, 17 (2002), 825–841. https://doi.org/10.1006/nimg.2002.1132 doi: 10.1006/nimg.2002.1132
[22]	H. Lee, K. Jung, D. W. Kang, N. Kim, Fully automated and real-time volumetric measurement of infarct core and penumbra in diffusion-and perfusion-weighted MRI of patients with hyper-acute stroke, J. Digit. Imaging, 33 (2020), 262–272. https://doi.org/10.1007/s10278-019-00222-2 doi: 10.1007/s10278-019-00222-2
[23]	S. M. Smith, Fast robust automated brain extraction, Hum. Brain Mapp., 17 (2002), 143–155. https://doi.org/10.1002/hbm.10062 doi: 10.1002/hbm.10062
[24]	M. Jenkinson, M. Pechaud, S. Smith, BET2: MR-based estimation of brain, skull and scalp surfaces, in Eleventh Annual Meeting of the Organization for Human Brain Mapping, 17 (2005), 167.
[25]	J. J. M. Van Griethuysen, A. Fedorov, C. Parmar, A. Hosny, N. Aucoin, V. Narayan, et al., Computational radiomics system to decode the radiographic phenotype, Cancer Res., 77 (2017), e104–e107. https://doi.org/10.1158/0008-5472.CAN-17-0339 doi: 10.1158/0008-5472.CAN-17-0339
[26]	Y. Zhang, B. Zhang, F. Liang, S. Liang, Y. Zhang, P. Yan, et al., Radiomics features on non-contrast-enhanced CT scan can precisely classify AVM-related hematomas from other spontaneous intraparenchymal hematoma types, Eur. Radiol., 29 (2019), 2157–2165. https://doi.org/10.1007/s00330-018-5747-x doi: 10.1007/s00330-018-5747-x
[27]	Muthukrishnan, R. and R. Rohini. LASSO: A feature selection technique in predictive modeling for machine learning, in 2016 IEEE International Conference on Advances in Computer Applications (ICACA), (2016), 18–20. https://doi.org/10.1109/ICACA.2016.7887916
[28]	X. Wu, H. Wang, F. Chen, L. Jin, J. Li, Y. Feng, et al., Rat model of reperfused partial liver infarction: characterization with multiparametric magnetic resonance imaging, microangiography, and histomorphology, Acta Radiol., 50 (2009), 276–287. https://doi.org/10.1080/02841850802647021 doi: 10.1080/02841850802647021
[29]	C. Wang, P. Miao, J. Liu, Z. Li, Y. Wei, Y. Wang, et al., Validation of cerebral blood flow connectivity as imaging prognostic biomarker on subcortical stroke, J. Neurochem., 159 (2021), 172–184. https://doi.org/10.1111/jnc.15359 doi: 10.1111/jnc.15359
[30]	D. A. Hernandez, R. P. H. Bokkers, R. V. Mirasol, M. Luby, E. C. Henning, J. G. Merino, et al., Pseudocontinuous arterial spin labeling quantifies relative cerebral blood flow in acute stroke, Stroke, 43 (2012), 753–758. https://doi.org/10.1161/STROKEAHA.111.635979 doi: 10.1161/STROKEAHA.111.635979
[31]	C. Wang, P. Miao, J. Liu, S. Wei, Y. Guo, Z. Li, et al., Cerebral blood flow features in chronic subcortical stroke: Lesion location-dependent study, Brain Res., 1706 (2019), 177–183. https://doi.org/10.1016/j.brainres.2018.11.009 doi: 10.1016/j.brainres.2018.11.009
[32]	T. Love, D. Swinney, E. Wong, R. Buxton, Perfusion imaging and stroke: A more sensitive measure of the brain bases of cognitive deficits, Aphasiology, 16 (2002), 873–883. https://doi.org/10.1080/02687030244000356 doi: 10.1080/02687030244000356
[33]	M. H. Lev, A. Z. Segal, J. Farkas, S. T. Hossain, C. Putman, G. J. Hunter, et al., Utility of perfusion-weighted CT imaging in acute middle cerebral artery stroke treated with intra-arterial thrombolysis-Prediction of final infarct volume and clinical outcome, Stroke, 32 (2001), 2021–2027. https://doi.org/10.1161/hs0901.095680 doi: 10.1161/hs0901.095680
[34]	C. Grefkes, G. R. Fink, Connectivity-based approaches in stroke and recovery of function, Lancet Neurol., 13 (2014), 206–216. https://doi.org/10.1016/S1474-4422(13)70264-3 doi: 10.1016/S1474-4422(13)70264-3
[35]	M Giacalone, P Rasti, N Debs, C Frindel, TH Cho, E. Grenier, Local spatio-temporal encoding of raw perfusion MRI for the prediction of final lesion in stroke, Med. Image Anal., 50 (2018), 117–126. https://doi.org/10.1016/j.media.2018.08.008 doi: 10.1016/j.media.2018.08.008
[36]	J. D. Jordan, W. J. Powers, Cerebral autoregulation and acute ischemic stroke, Am. J. Hypertens., 25 (2012), 946–950. https://doi.org/10.1038/ajh.2012.53 doi: 10.1038/ajh.2012.53
[37]	X. Yao, L. Mao, S. Lv, Z. Ren, W. Li, K. Ren, CT radiomics features as a diagnostic tool for classifying basal ganglia infarction onset time, J. Neurol. Sci., 412 (2020), 116730. https://doi.org/10.1016/j.jns.2020.116730 doi: 10.1016/j.jns.2020.116730
[38]	Z. Yi, L. Long, Y. Zeng, Z. Liu, Current advances and challenges in radiomics of brain tumors, Front. Oncol., 11 (2021). https://doi.org/10.3389/fonc.2021.732196 doi: 10.3389/fonc.2021.732196
[39]	Y. Zhang, B. Zhang, F. Liang, S. Liang, Y. Zhang, P. Yan, et al., Radiomics features on non-contrast-enhanced CT scan can precisely classify AVM-related hematomas from other spontaneous intraparenchymal hematoma types, Eur. Radiol., 29 (2019), 2157–2165. https://doi.org/10.1007/s00330-018-5747-x doi: 10.1007/s00330-018-5747-x
[40]	M. E. Mayerhoefer, A. Materka, G. Langs, I. Häggström, P. Szczypiński, P. Gibbs, et al., Introduction to radiomics, J. Nucl. Med., 61 (2020), 488–495. https://doi.org/10.2967/jnumed.118.222893 doi: 10.2967/jnumed.118.222893
[41]	M. Zhou, J. Scott, B. Chaudhury, L. Hall, D. Goldgof, K. W. Yeom, et al., Radiomics in brain tumor: Image assessment, quantitative feature descriptors, and machine-learning approaches, Am. J. Neuroradiol., 39 (2018), 208–216.

Reader Comments

Your name:*

Email:*
© 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)