Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2016 Apr 14;11(4):e0153369.
doi: 10.1371/journal.pone.0153369. eCollection 2016.

Cerebral Glioma Grading Using Bayesian Network with Features Extracted from Multiple Modalities of Magnetic Resonance Imaging

Jisu Hu  1   2 , Wenbo Wu  3 , Bin Zhu  3 , Huiting Wang  3 , Renyuan Liu  3 , Xin Zhang  3 , Ming Li  3 , Yongbo Yang  4 , Jing Yan  5 , Fengnan Niu  6 , Chuanshuai Tian  3 , Kun Wang  3 , Haiping Yu  3 , Weibo Chen  7 , Suiren Wan  1 , Yu Sun  1 , Bing Zhang  3
Affiliations

Cerebral Glioma Grading Using Bayesian Network with Features Extracted from Multiple Modalities of Magnetic Resonance Imaging

Jisu Hu et al. PLoS One. .

Erratum in

Abstract

Many modalities of magnetic resonance imaging (MRI) have been confirmed to be of great diagnostic value in glioma grading. Contrast enhanced T1-weighted imaging allows the recognition of blood-brain barrier breakdown. Perfusion weighted imaging and MR spectroscopic imaging enable the quantitative measurement of perfusion parameters and metabolic alterations respectively. These modalities can potentially improve the grading process in glioma if combined properly. In this study, Bayesian Network, which is a powerful and flexible method for probabilistic analysis under uncertainty, is used to combine features extracted from contrast enhanced T1-weighted imaging, perfusion weighted imaging and MR spectroscopic imaging. The networks were constructed using K2 algorithm along with manual determination and distribution parameters learned using maximum likelihood estimation. The grading performance was evaluated in a leave-one-out analysis, achieving an overall grading accuracy of 92.86% and an area under the curve of 0.9577 in the receiver operating characteristic analysis given all available features observed in the total 56 patients. Results and discussions show that Bayesian Network is promising in combining features from multiple modalities of MRI for improved grading performance.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: Author Weibo Chen is employed by Philips Healthcare. There are no patents, products in development or marketed products to declare. This does not alter the authors' adherence to all the PLoS ONE policies on sharing data and materials.

Figures

Fig 1
Fig 1. Examples of different enhancement extent.
The left column is a case of glioblastoma with apparent enhancement, the middle a case of high-grade glioma with slight enhancement (white arrow) and the right a low-grade glioma with negative enhancement. Upper row axial T2-weighted images. Lower row axial T1W+C images.
Fig 2
Fig 2. The pseudo-color maps of metabolite ratios in MRSI.
(A) Cho/Cr map. (B) NAA/Cr. (C) Lac/Cr map/ (D) Lip13/Cr map. (E) The VOI boundary (green) and the manually selected region (red).
Fig 3
Fig 3. Some BNs constructed in this study.
(A) The BN of perfusion features with nTTP. (B) The BN of perfusion features without nTTP. (C) The BN of T1W+C and perfusion features. (D) The BN of MRSI features with Lac/Cr. (E) The BN of MRSI features without Lac/Cr. (F) The final BN structure.
Fig 4
Fig 4. The ROC curves in the three datasets.
(A) The ROC curve in the 56 cases. (B) The ROC curve in the 51 cases. (C) The ROC curve in the 26 cases.
See this image and copyright information in PMC

References

    1. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, Jouvet A, et al. The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathologica. 2007;114(2):97–109. 10.1007/s00401-007-0243-4 WOS:000248046800001. - DOI - PMC - PubMed
    1. Law M, Yang S, Wang H, Babb JS, Johnson G, Cha S, et al. Glioma grading: Sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging. American Journal of Neuroradiology. 2003;24(10):1989–98. WOS:000186744800013. - PMC - PubMed
    1. Hakyemez B, Erdogan C, Ercan I, Ergin N, Uysal S, Atahan S. High-grade and low-grade gliomas: differentiation by using perfusion MR imaging. Clinical Radiology. 2005;60(4):493–502. 10.1016/j.crad.2004.09.009 WOS:000228259900009. - DOI - PubMed
    1. Barbier EL, Lamalle L, Decorps M. Methodology of brain perfusion imaging. Journal of Magnetic Resonance Imaging. 2001;13(4):496–520. 10.1002/jmri.1073 WOS:000171296200003. - DOI - PubMed
    1. Ostergaard L. Principles of cerebral perfusion imaging by bolus tracking. Journal of Magnetic Resonance Imaging. 2005;22(6):710–7. 10.1002/jmri.20460 WOS:000233767200006. - DOI - PubMed

Publication types