Share：
Share this content in WeChat
X
[Chinese][PDF]752170
• Experience Exchanges •
Prediction of Ki-67 expression level in glioblastoma by radiomics model based on T2WI
ZHU Xuechao  HE Yulin  WU Yingying  LI Bin  WAN Tianyi  TANG Xin  YU Qiuyue 

Cite this article as: Citation:Zhu XC, He YL, Wu YY, et al. Prediction of Ki-67 expression level in glioblastoma by radiomics model based on T2WI[J]. Chin J Magn Reson Imaging, 2021, 12(9): 53-56. DOI:10.12015/issn.1674-8034.2021.09.012.


[Abstract] Objective To explore the value of T2WI imaging model in predicting Ki-67 proliferation level of glioblastoma before operation. Materials andMethods The preoperative MRI images of 96 patients with glioblastoma diagnosed by pathology in our hospital were retrospectively analyzed. According to the expression level of Ki-67, the patients were divided into two groups: low expression group (Ki-67＜50%) and high expression group (Ki-67≥50). The volume of interest (VOI) was manually sketched on the T2WI axial image and the imaging features were extracted. All cases are divided into training group and test group according to 70%∶30%. The training group was used for feature screening and the establishment of machine learning models. Feature selection was completed by t-test and LASSO. After feature screening, three machine learning models of random forest (RF), Logistic regression and support vector machine (SVM) were established. The test group was used to verify the established model and draw ROC curves, and the results were expressed as accuracy, sensitivity, specificity and AUC.Results There was no significant difference in age and sex between Ki-67 low expression group and high expression group. Among the three machine learning models, the RF model has the highest diagnostic efficiency, and the accuracy, sensitivity, specificity and AUC are 0.72, 0.67, 0.76 and 0.72 respectively. The comprehensive diagnostic efficiency of the SVM model is the lowest, while the SVM model is between them.Conclusions The imaging model based on T2WI image has a certain value in predicting the level of Ki-67 expression in glioblastoma before operation, among which RF model is the best.
[Keywords] magnetic resonance imaging；glioblastoma；Ki-67；machine learning；radiomics；T2-weighted imaging

ZHU Xuechao   HE Yulin*   WU Yingying   LI Bin   WAN Tianyi   TANG Xin   YU Qiuyue  

Department of Radiology, the First Affiliated Hospital of Nanchang University, Nanchang 330006, China

HE YL, E-mail: 173386424@qq.com

Conflicts of interest   None.

Received  2021-02-09
Accepted  2021-04-19
DOI: 10.12015/issn.1674-8034.2021.09.012
Cite this article as: Citation:Zhu XC, He YL, Wu YY, et al. Prediction of Ki-67 expression level in glioblastoma by radiomics model based on T2WI[J]. Chin J Magn Reson Imaging, 2021, 12(9): 53-56. DOI:10.12015/issn.1674-8034.2021.09.012.

[1]
Blanchet L, Krooshof PWT, Postma GJ, et al. Discrimination between metastasis and glioblastoma multiforme based on morphometric analysis of mr images[J]. Am J Neuroradiol, 2011, 32(1): 67-73. DOI: 10.3174/ajnr.A2269.
[2]
Scholzen T, Gerdes J. The Ki-67 protein: From the known and the unknown[J]. J Cell Physiol, 2000, 182(3): 311-322. DOI: 10.1002/(SICI)1097-4652(200003)182:3<311::AID-JCP1>3.0.CO;2-9.
[3]
Wang X, Wang D, Yao Z, et al. Machine learning models for multiparametric glioma grading with quantitative result interpretations[J]. Front Neurosci, 2019, 12: 1046. DOI: 10.3389/fnins.2018.01046.
[4]
Sun BX, Zhang H, Zhang XX, et al. Advances in magnetic resonance imaging in the diagnosis of prostate cancer[J]. Chin J Magn Reson Imaging, 2019, 10(12): 947-950. DOI: 10.12015/issn.1674-8034.2019.12.017.
[5]
Su C, Jiang J, Zhang S, et al. Radiomics based on multicontrast MRI can precisely differentiate among glioma subtypes and predict tumour-proliferative behaviour[J]. Eur Radiol, 2019, 29(4): 1986-1996. DOI: 10.1007/s00330-018-5704-8.
[6]
Kumar V, Gu Y, Basu S, et al. Radiomics: the process and the challenges[J]. Magn Reson Imaging, 2012, 30(9): 1234-1248. DOI: 10.1016/j.mri.2012.06.010.
[7]
Limkin EJ, Sun R, Dercle L, et al. Promises and challenges for the implementation of computational medical imaging (radiomics) in oncology[J]. Ann Oncol, 2017, 28(6): 1191-1206. DOI: 10.1093/annonc/mdx034.
[8]
Beesley MF, McLaren KM. Cytokeratin 19 and galectin: 3 immunohistochemistry in the differential diagnosis of solitary thyroid nodules[J]. Histopathology, 2002, 41(3): 236-243. DOI: 10.1046/j.1365-2559.2002.01442.x.
[9]
Li J, Liu S, Qin Y, et al. High-order radiomics features based on T2 FLAIR MRI predict multiple glioma immunohistochemical features: a more precise and personalized gliomas management[J]. PLoS One, 2020,15(1): e227703. DOI: 10.1371/journal.pone.0227703.
[10]
Sun Y, Li H, Wu Y, et al. Correlations of mutations of IDH1,IDH2,p53 gene and Ki-67 protein expression in gliomas with the clinical features of pathological grading[J]. J Pract Med, 2018, 34(9): 1455-1459. DOI: 10.3969/j.issn.1006-5725.2018.09.012.
[11]
Xuan YH, Choi YL, Shin YK, et al. An immunohistochemical study of the expression of cell-cycle-regulated proteins p53, cyclin D1, RB, p27, Ki67 and MSH2 in gallbladder carcinoma and its precursor lesions[J]. Histol Histopathol, 2005, 20(1): 59. DOI: 10.14670/HH-20.59.
[12]
Swanton C. Intratumor heterogeneity: evolution through space and time[J]. Cancer Res, 2012, 72(19): 4875-4882. DOI: 10.1158/0008-5472.CAN-12-2217.
[13]
Cattoretti G, Becker MHG, Key G, et al. Monoclonal antibodies against recombinant parts of the Ki‐67 antigen (MIB 1 and MIB 3) detect proliferating cells in microwave‐processed formalin‐fixed paraffin sections[J]. J Pathol, 1992, 168(4): 357-363. DOI: 10.1002/path.1711680404.
[14]
Liang C, Cheng Z, Huang Y, et al. An MRI-based radiomics classifier for preoperative prediction of Ki-67 status in breast cancer[J]. Acad Radiol, 2018, 25(9): 1111-1117. DOI: 10.1016/j.acra.2018.01.006.
[15]
Gu Q, Feng Z, Liang Q, et al. Machine learning-based radiomics strategy for prediction of cell proliferation in non-small cell lung cancer[J]. Eur J Radiol, 2019, 118: 32-37. DOI: 10.1016/j.ejrad.2019.06.025.
[16]
Wu H, Han X, Wang Z, et al. Prediction of the Ki-67 marker index in hepatocellular carcinoma based on CT radiomics features[J]. Phys Med Biol, 2020,65(23): 235048. DOI: 10.1088/1361-6560/abac9c.
[17]
Park EK, Lee K, Seo BK, et al. Machine learning approaches to radiogenomics of breast cancer using low-dose perfusion computed tomography: predicting prognostic biomarkers and molecular subtypes[J]. Sci Report, 2019, 9(1): 17847. DOI: 10.1038/s41598-019-54371-z.
[18]
Gao M, Huang S, Pan X, et al. Machine learning-based radiomics predicting tumor grades and expression of multiple pathologic biomarkers in gliomas[J]. Front Oncol, 2020, 10: 1676. DOI: 10.3389/fonc.2020.01676.
[19]
Liu Z, Liu XT,Yin CC, et al. Application of T2WI machine learning in the identification of high-grade[J]. J Xi'an Jiaotong University (Med Sci), 2019, 40(5): 794-799. DOI: 10.7652/jdyxb201905025.
[20]
Ma W, Ji Y, Qi L, et al. Breast cancer Ki67 expression prediction by DCE-MRI radiomics features[J]. Clin Radiol, 2018, 73(10): 901-909. DOI: 10.1016/j.crad.2018.05.027.
[21]
Schieda N, Coffey N, Gulavita P, et al. Prostatic ductal adenocarcinoma: an aggressive tumour variant unrecognized on T2 weighted magnetic resonance imaging (MRI)[J]. Eur Radiol, 2014, 24(6): 1349-1356. DOI: 10.1007/s00330-014-3150-9.
[22]
Jha RC, Zanello PA, Nguyen XM, et al. Small hepatocellular carcinoma[J]. Acad Radiol, 2014, 21(11): 1455-1464. DOI: 10.1016/j.acra.2014.06.011.

PREV Evalution on the visualization of the extracranial branch of trigeminal nerve using the 3.0 T MR CE 3D-STIR TSE sequence
NEXT The correlation between diffusion kurtosis imaging and pathological condition in clear cell renal carcinoma
  


LINK


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