Share：
Share this content in WeChat
X
[Chinese][PDF]3092
• Clinical Article •
Multi-sequence MRI-based convolutional neural network predicts the methylation status of MGMT promoter in glioma
CHEN Xiaohua  ZHANG Ruodi  ZHOU Yunshu  LIU Shili  WANG Zhuo  ZHANG Shaoru  CHEN Zhiqiang 

CHEN X H, ZHANG R D, ZHOU Y S, et al. Multi-sequence MRI-based convolutional neural network predicts the methylation status of MGMT promoter in glioma[J]. Chin J Magn Reson Imaging, 2023, 14(8): 34-39, 78. DOI:10.12015/issn.1674-8034.2023.08.005.


[Abstract] Objective To investigate the value of a convolutional neural network model based on multi-sequence MRI to predict the promoter methylation status of O6-methylguanine-DNA-methyltransferase (MGMT) in glioma.Materials and Methods Retrospective analysis of clinical and MRI data of 161 patients with glioma confirmed by surgical pathology from November 2015 to June 2022 at Ningxia Medical University General Hospital, including 80 cases of MGMT promoter methylation type and 81 cases of unmethylated type. T2WI, T2 fluid-attenuated inversion recovery (T2-FLAIR) and contrast enhanced T1‐weighted imaging (CE-T1WI) of preoperative MRI were collected, and regions of interest (ROI) were outlined after preprocessing of all images. The images were randomly divided into training and validation sets according to 7∶3 after labeling. A 34-layer-residual neural network (ResNet34) was used to build T2WI, T2-FLAIR, enhanced T1WI and multiple sequence fusion models T2-net, T2f-net, TC-net and TS-net, respectively, to predict the methylation status of MGMT promoters. The area under the receiver operating characteristic (AUROC), area under the precision-recall curve (AUPRC), accuracy, specificity and sensitivity were used to assess model efficacy, and the predictive power was compared between models by DeLong test.Results All four prediction models T2-net, T2f-net, TC-net, and TS-net had good prediction efficacy, and the AUROC values of TS-net were higher than those of T2-net, T2f-net, and TC-net (training set: 0.930 vs. 0.859, 0.877, 0.920; validation set: 0.910 vs. 0.812, 0.840, 0.854). The AUPRC values of TS-net were higher than those of T2-net, T2f-net, and TC-net (training set: 0.912 vs. 0.860, 0.864, 0.908; validation set: 0.896 vs. 0.796, 0.826, 0.839). The AUROC values of TS-net in the validation set were all higher than those of T2-net, T2f-net, and TC-net, and the differences were all statistically significant. In addition, the differences in the training set were statistically significant compared with T2-net and T2f-net (DeLong test, P＜0.05).Conclusions Convolutional neural network models based on multi-sequence MRI fusion can accurately and non-invasively predict the MGMT methylation status of glioma, which is superior to single-sequence models and provides a reliable basis for guiding clinical treatment decisions and assessing the prognosis of glioma patients.
[Keywords] adult diffuse glioma；magnetic resonance imaging；deep learning；convolution neural network；O6-methylguanine-DNA-methyltransferase；molecular subtype

CHEN Xiaohua1   ZHANG Ruodi1   ZHOU Yunshu1   LIU Shili1   WANG Zhuo1   ZHANG Shaoru1   CHEN Zhiqiang1, 2, 3*  

1 Department of Radiology, General Hospital of Ningxia Medical University, Yinchuan 750004, China

2 Department of Radiology, the First Hospital Affiliated to Hainan Medical College, Haikou 570102, China

3 College of Clinical Medicine, Ningxia Medical University, Yinchuan 750004, China

Corresponding author: Chen ZQ, E-mail: zhiqiang_chen99@163.com

Conflicts of interest   None.

ACKNOWLEDGMENTS Key R & D Program of Ningxia Hui Autonomous Region (No. 2019BEG03033); Natural Science Foundation of Ningxia (No. 2022AAC03472); "Chunhui Project" of the Ministry of Education (No. Z2012002).
Received  2023-01-05
Accepted  2023-07-27
DOI: 10.12015/issn.1674-8034.2023.08.005
CHEN X H, ZHANG R D, ZHOU Y S, et al. Multi-sequence MRI-based convolutional neural network predicts the methylation status of MGMT promoter in glioma[J]. Chin J Magn Reson Imaging, 2023, 14(8): 34-39, 78. DOI:10.12015/issn.1674-8034.2023.08.005.

[1]
SLEDZINSKA P, BEBYN M G, FURTAK J, et al. Prognostic and predictive biomarkers in gliomas[J/OL]. Int J Mol Sci, 2021, 22(19): 10373 [2023-01-02]. https://www.mdpi.com/journal/ijms. DOI: 10.3390/ijms221910373.
[2]
MOLINARO A M, TAYLOR J W, WIENCKE J K, et al. Genetic and molecular epidemiology of adult diffuse glioma[J]. Nat Rev Neurol, 2019, 15(7): 405-417. DOI: 10.1038/s41582-019-0220-2.
[3]
WANG K, ZHANG S, SHI L, et al. The 2016 World Health Organization classification of tumors of the central nervous system: A summary[J]. Chin J Magn Reson Imaging, 2016, 7(12): 881-896. DOI: 10.12015/issn.1674-8034.2016.12.001.
[4]
LOUIS D N, PERRY A, WESSELING P, et al. The 2021 WHO Classification of Tumors of the Central Nervous System: a summary[J]. Neuro-Oncology, 2021, 23(8): 1231-1251. DOI: 10.1093/neuonc/noab106.
[5]
WELLER M, TABATABAI G, KASTNER B, et al. MGMT promoter methylation is a strong prognostic biomarker for benefit from dose-intensified temozolomide rechallenge in progressive glioblastoma: The director trial[J]. Clin Cancer Res, 2015, 21(9): 2057-2064. DOI: 10.1158/1078-0432.CCR-14-2737.
[6]
ZHANG K, WANG X Q, ZHOU B, et al. The prognostic value of MGMT promoter methylation in Glioblastoma multiforme: a meta-analysis[J]. Fam Cancer, 2013, 12(3): 449-458. DOI: 10.1007/s10689-013-9607-1.
[7]
LI H, LI J, CHENG G, et al. IDH mutation and MGMT promoter methylation are associated with the pseudoprogression and improved prognosis of glioblastoma multiforme patients who have undergone concurrent and adjuvant temozolomide-based chemoradiotherapy[J]. Clin Neurol Neurosurg, 2016, 151: 31-36. DOI: 10.1016/j.clineuro.2016.10.004.
[8]
SMITS M, VAN DEN BENT M J. Imaging correlates of adult glioma genotypes[J]. Radiology, 2017, 284(2): 316-331. DOI: 10.1148/radiol.2017151930.
[9]
YU X, WU Y P, BAI Y, et al. Application of 3D convolutional neural network based on fusion of multiple sequence MRI to evaluate the survival prediction of patients with glioma[J]. Chin J Magn Reson Imaging, 2023, 14(3: 12-16. DOI: 10.12015/issn.1674-8034.2023.03.003.
[10]
LIU B, ZENG Q, HUANG J, et al. IVIM using convolutional neural networks predicts microvascular invasion in HCC[J]. Eur Radiol, 2022, 32(10: 7185-7195. DOI: 10.1007/s00330-022-08927-9.
[11]
TRAN K A, KONDRASHOVA O, BRADLEY A, et al. Deep learning in cancer diagnosis, prognosis and treatment selection[J/OL]. Genome Med, 2021, 13(1: 152 [2023-01-02]. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8477474/. DOI: 10.1186/s13073-021-00968-x.
[12]
ECHLE A, RINDTORFF N T, BRINKER T J, et al. Deep learning in cancer pathology: a new generation of clinical biomarkers[J]. Br J Cancer, 2021, 124(4: 686-696. DOI: 10.1038/s41416-020-01122-x.
[13]
RAZEK A A K A, ALKSAS A, SHEHATA M, et al. Clinical applications of artificial intelligence and radiomics in neuro-oncology imaging[J/OL]. Insights Imaging, 2021, 12(1: 152 [2023-01-02]. https://insightsimaging.springeropen.com/articles/10.1186/s13244-021-01102-6. DOI: 10.1186/s13244-021-01102-6.
[14]
RUDIE J D, RAUSCHECKER A M, BRYAN R N, et al. Emerging applications of artificial intelligence in neuro-oncology[J]. Radiology, 2019, 290(3: 607-618. DOI: 10.1148/radiol.2018181928.
[15]
ANWAR S M, MAJID M, QAYYUM A, et al. Medical image analysis using convolutional neural networks: a review[J/OL]. J Med Syst, 2018, 42(11: 226 [2023-01-02]. https://doi.org/10.1007/s10916-018-1088-1. DOI: 10.1007/s10916-018-1088-1.
[16]
CHEN X, ZENG M, TONG Y C, et al. Automatic prediction of MGMT status in glioblastoma via deep learning-based MR image analysis[J/OL]. Biomed Res Int, 2020, 2020: 9258649 [2023-01-02]. https://www.hindawi.com/journals/bmri/2020/9258649/. DOI: 10.1155/2020/9258649.
[17]
TUSTISON N J, AVANTS B B, COOK P A, et al. N4ITK: improved N3 bias correction[J]. IEEE Trans Med Imaging, 2010, 29(6: 1310-1320. DOI: 10.1109/TMI.2010.2046908.
[18]
HE K M, ZHANG X Y, REN S Q, et al. Identity Mappings in Deep Residual Networks[M]. In: Lecture Notes in Computer Science, 2016, 9908: 630-645 [2023-01-02]. https://link.springer.com/chapter/10.1007/978-3-319-46493-0_38. DOI: 10.1007/978-3-319-46493-0_38.
[19]
MATSUI Y, MARUYAMA T, NITTA M, et al. Prediction of lower-grade glioma molecular subtypes using deep learning[J]. J Neurooncol, 2020, 146(2: 321-327. DOI: 10.1007/s11060-019-03376-9.
[20]
YOON R G, KIM H S, PAIK W, et al. Different diagnostic values of imaging parameters to predict pseudoprogression in glioblastoma subgroups stratified by MGMT promoter methylation[J]. Eur Radiol, 2017, 27(1: 255-266. DOI: 10.1007/s00330-016-4346-y.
[21]
GE S N, SHI Y W, ZHU G, et al. Molecular pathological markers correlated with the recurrence patterns of glioma[J/OL]. Front Oncol, 2020, 10: 565045 [2023-01-02]. https://www.frontiersin.org/articles/10.3389/fonc.2020.565045/full. DOI: 10.3389/fonc.2020.565045.
[22]
FAN Z W, LIU Y K, LI S W, et al. Association of tumor growth rates with molecular biomarker status: a longitudinal study of high-grade glioma[J]. Aging (Albany NY), 2020, 12(9: 7908-7926. DOI: 10.18632/aging.103110.
[23]
YABE M, MIRANDA R N, MEDEIROS L J. Hepatosplenic T-cell Lymphoma: a review of clinicopathologic features, pathogenesis, and prognostic factors[J]. Hum Pathol, 2018, 74: 5-16. DOI: 10.1016/j.humpath.2018.01.005.
[24]
YUAN P X, GAO Y, WU Q, et al. The value of DCE and MAP-MRI in predicting the methylation status of MGMT promoter in high-grade glioma[J]. Chin J Magn Reson Imaging, 2023, 14(5: 85-91. DOI: 10.12015/issn.1674-8034.2023.05.016.
[25]
AHN S S, SHIN N Y, CHANG J H, et al. Prediction of methylguanine methyltransferase promoter methylation in glioblastoma using dynamic contrast-enhanced magnetic resonance and diffusion tensor imaging[J]. J Neurosurg, 2014, 121(2: 367-373. DOI: 10.3171/2014.5.JNS132279.
[26]
MOON W J, CHOI J W, ROH H G, et al. Imaging parameters of high grade gliomas in relation to the MGMT promoter methylation status: the CT, diffusion tensor imaging, and perfusion MR imaging[J]. Neuroradiology, 2012, 54(6: 555-563. DOI: 10.1007/s00234-011-0947-y.
[27]
ESTEVA A, KUPREL B, NOVOA R A, et al. Dermatologist-level classification of skin cancer with deep neural networks[J]. Nature, 2017, 19(11-12: 407-408. DOI: 10.1038/nature21056.
[28]
ZHAO H M, ZHANG H. Research progress of intelligent image prediction of MGMT methylation status in high-grade glioma[J]. Chin J Magn Reson Imaging, 2022, 13(2: 130-132. DOI: 10.12015/issn.1674-8034.2022.02.032.
[29]
LI Y M, WEI D, LIU X, et al. Molecular subtyping of diffuse gliomas using magnetic resonance imaging: comparison and correlation between radiomics and deep learning[J]. Eur Radiol, 2022, 32(2: 747-758. DOI: 10.1007/s00330-021-08237-6.
[30]
LI D R, HU W J, LIU G Y, et al. The MRI-based 3D-ResNet101 deep learning model for predicting preoperative grading of gliomas: A multicenter study[J]. Chin J Magn Reson Imaging, 2023, 14(5: 25-30. DOI: 10.12015/issn.1674-8034.2023.05.006.
[31]
CHOI Y S, BAE S, CHANG J H, et al. Fully automated hybrid approach to predict the IDH mutation status of gliomas via deep learning and radiomics[J]. Neuro Oncol, 2021, 23(2: 304-313. DOI: 10.1093/neuonc/noaa177.
[32]
CHANG K, BAI H X, ZHOU H, et al. Residual convolutional neural network for the determination of IDH status in low- and high-grade gliomas from MR imaging[J]. Clin Cancer Res, 2018, 24(5: 1073-1081. DOI: 10.1158/1078-0432.CCR-17-2236.
[33]
KORFIATIS P, KLINE T L, LACHANCE D H, et al. Residual deep convolutional neural network predicts mgmt methylation status[J]. J Digit Imaging, 2017, 30(5: 622-628. DOI: 10.1007/s10278-017-0009-z.
[34]
XUE C Q, DU X H, JIN L, et al. Prediction of methylation status of MGMT promoter in WHO grade Ⅱ, Ⅲ glioma based on MRI deep learning model[J]. Chin J Radiol, 2021, 55(7: 734-738. DOI: 10.3760/cma.j.cn112149-20200825-01029.
[35]
YOGANANDA C G B, SHAH B R, NALAWADE S S, et al. MRI-based deep-learning method for determining glioma mgmt promoter methylation status[J]. AJNR Am J Neuroradiol, 2021, 42(5: 845-852. DOI: 10.3174/ajnr.A7029.
[36]
DI IEVA A, RUSSO C, LIU S, et al. Application of deep learning for automatic segmentation of brain tumors on magnetic resonance imaging: a heuristic approach in theclinical scenario[J]. Neuroradiology, 2021, 63(8: 1253-1262. DOI: 10.1007/s00234-021-02649-3.

PREV Predicting IDH1 gene mutation of gliomas by combining clinical and imaging features with multiple sequence radiomics
NEXT Clinical and imaging features and prognostic factors of acute ischemic stroke in patients with Trousseau,s syndrome
  


LINK


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