Share：
Share this content in WeChat
X
[Chinese][PDF]96696
• Clinical Article •
Investigation of automated glioma grading using DCE-MRI quantitative parameters
NAN Hai-yan  ZHANG Xin  YAN Lin-feng  YANG Yang  HAN Yu  WANG Wen  CUI Guang-bin 

DOI:10.12015/issn.1674-8034.2018.07.003.


[Abstract] Objective: To develop a non-invasive and automated preoperative glioma grading system based on quantitative parameters derived from dynamic contrast-enhanced MRI (DCE-MRI) data.Materials and Methods: A total of 98 histologically confirmed glioma patients were recruited in this study, including 28 low-grade gliomas (LGGs) and 70 high-grade gliomas (HGGs), who underwent preoperative conventional MRI and DCE-MRI scans. Parametric maps such as AUCAIF, Ktrans, Kep, Ve and Vp were derived using the NordicICE software. Statistically histogram indices of the whole tumor region were calculated from each parametric map and formed the corresponding parametric feature dataset. Support vector machine recursive feature elimination (SVM-RFE) feature selection strategy and SVM classifier were jointly used to train the glioma grade discrimination models. The classification performance of each parametric model was evaluated by 10-fold cross-validation method.Results: The AUCAIF, Ktrans, Kep and Ve parametric classification model achieved relatively high accuracy and area under the curve (AUC) values (all of them ≥0.75), except for Vp. When all of the parametric features were combined, the accuracy and AUC values of the trained grading model both increased to 0.864, obviously higher than each independent parametric features.Conclusions: Utilizing the quantitative parameters provided by DCE-MRI and high-efficiency machine learning techniques, it is possible to establish a non-invasive and automated preoperative glioma grading model with high efficiency. This study will provide valuable reference for making treatment plans and increasing prognosis in the future.
[Keywords] Glioma；Neoplasm grading；Magnetic resonance imaging

NAN Hai-yan Department of Radiology, Tangdu Hospital, the Fourth Military Medical University, Xi’an 710038, China

ZHANG Xin Department of Radiology, Tangdu Hospital, the Fourth Military Medical University, Xi’an 710038, China

YAN Lin-feng Department of Radiology, Tangdu Hospital, the Fourth Military Medical University, Xi’an 710038, China

YANG Yang Department of Radiology, Tangdu Hospital, the Fourth Military Medical University, Xi’an 710038, China

HAN Yu Department of Radiology, Tangdu Hospital, the Fourth Military Medical University, Xi’an 710038, China

WANG Wen Department of Radiology, Tangdu Hospital, the Fourth Military Medical University, Xi’an 710038, China

CUI Guang-bin* Department of Radiology, Tangdu Hospital, the Fourth Military Medical University, Xi’an 710038, China

*Correspondence to: Cui GB, E-mail: cgbtd@126.com

Conflicts of interest   None.

ACKNOWLEDGMENTS  Supported by Social Development and Scientific Research Projects of Shaanxi Province NO. 2014JZ2-007 and Science and Innovation Development Fund of Tangdu Hospital of Air Force Military Medical University (the Fourth Military Medical University) NO. 2016LCYJ001
Received  2018-03-18
Accepted  2018-05-15
DOI: 10.12015/issn.1674-8034.2018.07.003
DOI:10.12015/issn.1674-8034.2018.07.003.

[1]
Togao O, Hiwatashi A, Yamashita K, et al. Differentiation of high-grade and low-grade diffuse gliomas by intravoxel incoherent motion MR imaging. Neuro Oncol, 2016, 18(1): 132-141.
[2]
Inano R, Oishi N, Kunieda T, et al. Voxel-based clustered imaging by multiparameter diffusion tensor images for glioma grading. Neuroimage Clin, 2014, 5: 396-407.
[3]
Sahoo P, Gupta PK, Awasthi A, et al. Comparison of actual with default hematocrit value in dynamic contrast enhanced MR perfusion quantification in grading of human glioma. Magn Reson Imaging, 2016, 34(8): 1071-1077.
[4]
Arevalo-Perez J, Peck KK, Young RJ, et al. Dynamic contrast-enhanced perfusion MRI and diffusion-weighted imaging in grading of gliomas. J Neuroimaging, 2015, 25(5): 792-798.
[5]
Falk A, Fahlstrom M, Rostrup E, et al. Discrimination between glioma grades Ⅱ and Ⅲ in suspected low-grade gliomas using dynamic contrast-enhanced and dynamic susceptibility contrast perfusion MR imaging: a histogram analysis approach. Neuroradiology, 2014, 56(12): 1031-1038.
[6]
Zhang N, Zhang L, Qiu B, et al. Correlation of volume transfer coefficient Ktrans with histopathologic grades of gliomas. J Magn Reson imaging, 2012, 36(2): 355-363.
[7]
张微，牛蕾，马敏阁. DCE-MRI在高、低级别脑胶质瘤及脑膜瘤中的鉴别诊断．磁共振成像, 2015, 6(8): 566-570.
[8]
贾龙威，牛蕾，马文帅．血流动力学双室模型Extended Tofts Linear在脑胶质瘤DCE-MRI渗透性定量分析的复测性及有效性研究．磁共振成像, 2015, 6(8): 571-574.
[9]
Choi HS, Kim AH, Ahn SS, et al. Glioma grading capability: comparisons among parameters from dynamic contrast-enhanced MRI and ADC value on DWI. Korean J Radiol. 2013, 14(3): 487-492.
[10]
Fehr D, Veeraraghavan H, Wibmer A, et al. Automatic classification of prostate cancer Gleason scores from multiparametric magnetic resonance images. Proc Natl Acad Sci U S A, 2015, 112(46): E6265-6273.
[11]
Macyszyn L, Akbari H, Pisapia JM, et al. Imaging patterns predict patient survival and molecular subtype in glioblastoma via machine learning techniques. Neuro Oncol, 2016, 18(3): 417-425.
[12]
Zollner FG, Emblem KE, Schad LR. SVM-based glioma grading: optimization by feature reduction analysis. Z Med Phys, 2012, 22(3): 205-214.
[13]
Bergamino M, Barletta L, Castellan L, et al. Dynamic contrast-enhanced MRI in the study of brain tumors. Comparison Between the Extended Tofts-Kety Model and a Phenomenological Universalities (PUN) Algorithm. J Digit Imaging, 2015, 28(6): 748-754.
[14]
Huang ML, Hung YH, Lee WM, et al. SVM-RFE based feature selection and Taguchi parameters optimization for multiclass SVM classifier. Scientific World Journal, 2014: 795624.
[15]
Kang Y, Choi SH, Kim YJ, et al. Gliomas: histogram analysis of apparent diffusion coefficient maps with standard- or high-b-value diffusion-weighted MR imaging--correlation with tumor grade. Radiology, 2011, 261(3): 882-890.
[16]
Ryu YJ, Choi SH, Park SJ, et al. Glioma: application of whole-tumor texture analysis of diffusion-weighted imaging for the evaluation of tumor heterogeneity. PloS One, 2014, 9(9): e108335.

PREV The effect of leucine on feeding center and reward system in T2DM patients and healthy control subjects: a resting-state fMRI study
NEXT Effects of KIBRA polymorphism on the volume of hippocampus subfield in healthy young volunteers
  


LINK


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