Malignant brain tumor classification using the random forest method

Lichi Zhang, Han Zhang, Islem Rekik, Yaozong Gao, Qian Wang, Dinggang Shen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Brain tumor grading is pivotal in treatment planning. Contrast-enhanced T1-weighted MR image is commonly used for grading. However, the classification of different types of high-grade gliomas using T1-weighted MR images is still challenging, due to the lack of imaging biomarkers. Previous studies only focused on simple visual features, ignoring rich information provided by MR images. In this paper, we propose an automatic classification pipeline using random forest to differentiate the WHO Grade III and Grade IV gliomas, by extracting discriminative features based on 3D patches. The proposed pipeline consists of three main steps in both the training and the testing stages. First, we select numerous 3D patches in and around the tumor regions of the given MR images. This can suppress the intensity information from the normal region, which is trivial for the classification process. Second, we extract features based on both patch-wise information and subject-wise clinical information, and then we refine this step to optimize the performance of malignant tumor classification. Third, we incorporate the classification forest for training/testing the classifier. We validate the proposed framework on 96 malignant brain tumor patients that consist of both Grade III (N = 38) and Grade IV gliomas (N = 58). The experiments show that the proposed framework has demonstrated its validity in the application of high-grade gliomas classification, which may help improve the poor prognosis of high-grade gliomas.

LanguageEnglish
Title of host publicationStructural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings
EditorsEdwin R. Hancock, Tin Kam Ho, Battista Biggio, Richard C. Wilson, Antonio Robles-Kelly, Xiao Bai
PublisherSpringer Verlag
Pages14-21
Number of pages8
ISBN (Print)9783319977843
DOIs
Publication statusPublished - 2 Aug 2018
EventJoint IAPR International Workshops on Structural and Syntactic Pattern Recognition, SSPR 2018 and Statistical Techniques in Pattern Recognition, SPR 2018 - Beijing, China
Duration: 17 Aug 201819 Aug 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11004 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceJoint IAPR International Workshops on Structural and Syntactic Pattern Recognition, SSPR 2018 and Statistical Techniques in Pattern Recognition, SPR 2018
CountryChina
CityBeijing
Period17/08/1819/08/18

Fingerprint

Brain Tumor
Random Forest
Tumors
Brain
Patch
Grading
Tumor
Pipelines
Testing
Prognosis
Biomarkers
Differentiate
Trivial
Classifiers
Classifier
Optimise
Imaging
Planning
Imaging techniques
Experiment

Cite this

Zhang, L., Zhang, H., Rekik, I., Gao, Y., Wang, Q., & Shen, D. (2018). Malignant brain tumor classification using the random forest method. In E. R. Hancock, T. K. Ho, B. Biggio, R. C. Wilson, A. Robles-Kelly, & X. Bai (Eds.), Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings (pp. 14-21). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11004 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-97785-0_2
Zhang, Lichi ; Zhang, Han ; Rekik, Islem ; Gao, Yaozong ; Wang, Qian ; Shen, Dinggang. / Malignant brain tumor classification using the random forest method. Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings. editor / Edwin R. Hancock ; Tin Kam Ho ; Battista Biggio ; Richard C. Wilson ; Antonio Robles-Kelly ; Xiao Bai. Springer Verlag, 2018. pp. 14-21 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{d2ee1b0e59da41dab57a139817c41262,
title = "Malignant brain tumor classification using the random forest method",
abstract = "Brain tumor grading is pivotal in treatment planning. Contrast-enhanced T1-weighted MR image is commonly used for grading. However, the classification of different types of high-grade gliomas using T1-weighted MR images is still challenging, due to the lack of imaging biomarkers. Previous studies only focused on simple visual features, ignoring rich information provided by MR images. In this paper, we propose an automatic classification pipeline using random forest to differentiate the WHO Grade III and Grade IV gliomas, by extracting discriminative features based on 3D patches. The proposed pipeline consists of three main steps in both the training and the testing stages. First, we select numerous 3D patches in and around the tumor regions of the given MR images. This can suppress the intensity information from the normal region, which is trivial for the classification process. Second, we extract features based on both patch-wise information and subject-wise clinical information, and then we refine this step to optimize the performance of malignant tumor classification. Third, we incorporate the classification forest for training/testing the classifier. We validate the proposed framework on 96 malignant brain tumor patients that consist of both Grade III (N = 38) and Grade IV gliomas (N = 58). The experiments show that the proposed framework has demonstrated its validity in the application of high-grade gliomas classification, which may help improve the poor prognosis of high-grade gliomas.",
author = "Lichi Zhang and Han Zhang and Islem Rekik and Yaozong Gao and Qian Wang and Dinggang Shen",
year = "2018",
month = "8",
day = "2",
doi = "10.1007/978-3-319-97785-0_2",
language = "English",
isbn = "9783319977843",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "14--21",
editor = "Hancock, {Edwin R.} and Ho, {Tin Kam} and Battista Biggio and Wilson, {Richard C.} and Antonio Robles-Kelly and Xiao Bai",
booktitle = "Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings",
address = "Germany",

}

Zhang, L, Zhang, H, Rekik, I, Gao, Y, Wang, Q & Shen, D 2018, Malignant brain tumor classification using the random forest method. in ER Hancock, TK Ho, B Biggio, RC Wilson, A Robles-Kelly & X Bai (eds), Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11004 LNCS, Springer Verlag, pp. 14-21, Joint IAPR International Workshops on Structural and Syntactic Pattern Recognition, SSPR 2018 and Statistical Techniques in Pattern Recognition, SPR 2018, Beijing, China, 17/08/18. https://doi.org/10.1007/978-3-319-97785-0_2

Malignant brain tumor classification using the random forest method. / Zhang, Lichi; Zhang, Han; Rekik, Islem; Gao, Yaozong; Wang, Qian; Shen, Dinggang.

Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings. ed. / Edwin R. Hancock; Tin Kam Ho; Battista Biggio; Richard C. Wilson; Antonio Robles-Kelly; Xiao Bai. Springer Verlag, 2018. p. 14-21 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11004 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Malignant brain tumor classification using the random forest method

AU - Zhang, Lichi

AU - Zhang, Han

AU - Rekik, Islem

AU - Gao, Yaozong

AU - Wang, Qian

AU - Shen, Dinggang

PY - 2018/8/2

Y1 - 2018/8/2

N2 - Brain tumor grading is pivotal in treatment planning. Contrast-enhanced T1-weighted MR image is commonly used for grading. However, the classification of different types of high-grade gliomas using T1-weighted MR images is still challenging, due to the lack of imaging biomarkers. Previous studies only focused on simple visual features, ignoring rich information provided by MR images. In this paper, we propose an automatic classification pipeline using random forest to differentiate the WHO Grade III and Grade IV gliomas, by extracting discriminative features based on 3D patches. The proposed pipeline consists of three main steps in both the training and the testing stages. First, we select numerous 3D patches in and around the tumor regions of the given MR images. This can suppress the intensity information from the normal region, which is trivial for the classification process. Second, we extract features based on both patch-wise information and subject-wise clinical information, and then we refine this step to optimize the performance of malignant tumor classification. Third, we incorporate the classification forest for training/testing the classifier. We validate the proposed framework on 96 malignant brain tumor patients that consist of both Grade III (N = 38) and Grade IV gliomas (N = 58). The experiments show that the proposed framework has demonstrated its validity in the application of high-grade gliomas classification, which may help improve the poor prognosis of high-grade gliomas.

AB - Brain tumor grading is pivotal in treatment planning. Contrast-enhanced T1-weighted MR image is commonly used for grading. However, the classification of different types of high-grade gliomas using T1-weighted MR images is still challenging, due to the lack of imaging biomarkers. Previous studies only focused on simple visual features, ignoring rich information provided by MR images. In this paper, we propose an automatic classification pipeline using random forest to differentiate the WHO Grade III and Grade IV gliomas, by extracting discriminative features based on 3D patches. The proposed pipeline consists of three main steps in both the training and the testing stages. First, we select numerous 3D patches in and around the tumor regions of the given MR images. This can suppress the intensity information from the normal region, which is trivial for the classification process. Second, we extract features based on both patch-wise information and subject-wise clinical information, and then we refine this step to optimize the performance of malignant tumor classification. Third, we incorporate the classification forest for training/testing the classifier. We validate the proposed framework on 96 malignant brain tumor patients that consist of both Grade III (N = 38) and Grade IV gliomas (N = 58). The experiments show that the proposed framework has demonstrated its validity in the application of high-grade gliomas classification, which may help improve the poor prognosis of high-grade gliomas.

UR - http://www.scopus.com/inward/record.url?scp=85052207828&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-97785-0_2

DO - 10.1007/978-3-319-97785-0_2

M3 - Conference contribution

SN - 9783319977843

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 14

EP - 21

BT - Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings

A2 - Hancock, Edwin R.

A2 - Ho, Tin Kam

A2 - Biggio, Battista

A2 - Wilson, Richard C.

A2 - Robles-Kelly, Antonio

A2 - Bai, Xiao

PB - Springer Verlag

ER -

Zhang L, Zhang H, Rekik I, Gao Y, Wang Q, Shen D. Malignant brain tumor classification using the random forest method. In Hancock ER, Ho TK, Biggio B, Wilson RC, Robles-Kelly A, Bai X, editors, Structural, Syntactic, and Statistical Pattern Recognition - Joint IAPR International Workshop, S+SSPR 2018, Proceedings. Springer Verlag. 2018. p. 14-21. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-97785-0_2