Imaging biomarker analysis of advanced multiparametric MRI for glioma grading

A Vamvakas; S Williams; K Theodorou; E Kapsalaki; K Fountas; C Kappas; K Vassiou; I Tsougos

doi:10.1016/j.ejmp.2019.03.014

Imaging biomarker analysis of advanced multiparametric MRI for glioma grading

A Vamvakas, S Williams, K Theodorou, E Kapsalaki, K Fountas, C Kappas, K Vassiou, I Tsougos

King's College London

Research output: Contribution to journal › Article › peer-review

86 Citations (Scopus)

Abstract

AIMS AND OBJECTIVES: To investigate the value of advanced multiparametric MR imaging biomarker analysis based on radiomic features and machine learning classification, in the non-invasive evaluation of tumor heterogeneity towards the differentiation of Low Grade vs. High Grade Gliomas.

METHODS AND MATERIALS: Forty histologically confirmed glioma patients (20 LGG and 20 HGG) who underwent a standard 3T-MRI tumor protocol with conventional (T1 pre/post-contrast, T2-FSE, T2-FLAIR) and advanced techniques (Diffusion Tensor and Perfusion Imaging, 1H-MR Spectroscopy), were included. A semi-automated segmentation technique, based on T1W-C and DTI, was used for tumor core delineation in all available parametric maps. 3D Texture analysis considered 12 Histogram, 11 Co-Occurrence Matrix (GLCM) and 5 Run Length Matrix (GLRLM) features, derived from p, q, MD, FA, T1W-C, T2W-FSE, T2W-FLAIR and raw DSCE data. Along with 1H-MRS metabolic ratios and mean rCBV values, a total of 581 attributes for each subject were obtained. A Support Vector Machine - Recursive Feature Elimination (SVM-RFE) algorithm and SVM classifier were utilized for feature selection and classification, respectively.

RESULTS: Three different SVM classifiers were evaluated with consecutively SVM-RFE feature subsets. Linear SMO classifier demonstrated the highest performance for determining the optimal feature subset. Finally, 21 SVM-RFE top-ranked features were adopted, for training and testing the SMO classifier with leave-one-out cross-validation, achieving 95.5% Accuracy, 95% Sensitivity, 96% Specificity and 95.5% Area Under ROC Curve.

CONCLUSION: Results demonstrate that quantitative analysis of phenotypic characteristics, based on advanced multiparametric MR neuroimaging data and texture features, utilizing state-of-the-art radiomic analysis methods, can significantly contribute to the pre-treatment glioma grade differentiation.

Original language	English
Pages (from-to)	188-198
Number of pages	11
Journal	Physica Medica
Volume	60
Early online date	23 Mar 2019
DOIs	https://doi.org/10.1016/j.ejmp.2019.03.014
Publication status	E-pub ahead of print - 23 Mar 2019

Keywords

Glioma grading
Machine learning
Multiparametric MRI
Radiomics

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1016/j.ejmp.2019.03.014

Cite this

@article{625385411859428da17740373896a77c,

title = "Imaging biomarker analysis of advanced multiparametric MRI for glioma grading",

abstract = "AIMS AND OBJECTIVES: To investigate the value of advanced multiparametric MR imaging biomarker analysis based on radiomic features and machine learning classification, in the non-invasive evaluation of tumor heterogeneity towards the differentiation of Low Grade vs. High Grade Gliomas.METHODS AND MATERIALS: Forty histologically confirmed glioma patients (20 LGG and 20 HGG) who underwent a standard 3T-MRI tumor protocol with conventional (T1 pre/post-contrast, T2-FSE, T2-FLAIR) and advanced techniques (Diffusion Tensor and Perfusion Imaging, 1H-MR Spectroscopy), were included. A semi-automated segmentation technique, based on T1W-C and DTI, was used for tumor core delineation in all available parametric maps. 3D Texture analysis considered 12 Histogram, 11 Co-Occurrence Matrix (GLCM) and 5 Run Length Matrix (GLRLM) features, derived from p, q, MD, FA, T1W-C, T2W-FSE, T2W-FLAIR and raw DSCE data. Along with 1H-MRS metabolic ratios and mean rCBV values, a total of 581 attributes for each subject were obtained. A Support Vector Machine - Recursive Feature Elimination (SVM-RFE) algorithm and SVM classifier were utilized for feature selection and classification, respectively.RESULTS: Three different SVM classifiers were evaluated with consecutively SVM-RFE feature subsets. Linear SMO classifier demonstrated the highest performance for determining the optimal feature subset. Finally, 21 SVM-RFE top-ranked features were adopted, for training and testing the SMO classifier with leave-one-out cross-validation, achieving 95.5% Accuracy, 95% Sensitivity, 96% Specificity and 95.5% Area Under ROC Curve.CONCLUSION: Results demonstrate that quantitative analysis of phenotypic characteristics, based on advanced multiparametric MR neuroimaging data and texture features, utilizing state-of-the-art radiomic analysis methods, can significantly contribute to the pre-treatment glioma grade differentiation.",

keywords = "Glioma grading, Machine learning, Multiparametric MRI, Radiomics",

author = "A Vamvakas and S Williams and K Theodorou and E Kapsalaki and K Fountas and C Kappas and K Vassiou and I Tsougos",

year = "2019",

month = mar,

day = "23",

doi = "10.1016/j.ejmp.2019.03.014",

language = "English",

volume = "60",

pages = "188--198",

journal = "Physica Medica",

issn = "1120-1797",

publisher = "Associazione Italiana di Fisica Medica",

}

TY - JOUR

T1 - Imaging biomarker analysis of advanced multiparametric MRI for glioma grading

AU - Vamvakas, A

AU - Williams, S

AU - Theodorou, K

AU - Kapsalaki, E

AU - Fountas, K

AU - Kappas, C

AU - Vassiou, K

AU - Tsougos, I

PY - 2019/3/23

Y1 - 2019/3/23

N2 - AIMS AND OBJECTIVES: To investigate the value of advanced multiparametric MR imaging biomarker analysis based on radiomic features and machine learning classification, in the non-invasive evaluation of tumor heterogeneity towards the differentiation of Low Grade vs. High Grade Gliomas.METHODS AND MATERIALS: Forty histologically confirmed glioma patients (20 LGG and 20 HGG) who underwent a standard 3T-MRI tumor protocol with conventional (T1 pre/post-contrast, T2-FSE, T2-FLAIR) and advanced techniques (Diffusion Tensor and Perfusion Imaging, 1H-MR Spectroscopy), were included. A semi-automated segmentation technique, based on T1W-C and DTI, was used for tumor core delineation in all available parametric maps. 3D Texture analysis considered 12 Histogram, 11 Co-Occurrence Matrix (GLCM) and 5 Run Length Matrix (GLRLM) features, derived from p, q, MD, FA, T1W-C, T2W-FSE, T2W-FLAIR and raw DSCE data. Along with 1H-MRS metabolic ratios and mean rCBV values, a total of 581 attributes for each subject were obtained. A Support Vector Machine - Recursive Feature Elimination (SVM-RFE) algorithm and SVM classifier were utilized for feature selection and classification, respectively.RESULTS: Three different SVM classifiers were evaluated with consecutively SVM-RFE feature subsets. Linear SMO classifier demonstrated the highest performance for determining the optimal feature subset. Finally, 21 SVM-RFE top-ranked features were adopted, for training and testing the SMO classifier with leave-one-out cross-validation, achieving 95.5% Accuracy, 95% Sensitivity, 96% Specificity and 95.5% Area Under ROC Curve.CONCLUSION: Results demonstrate that quantitative analysis of phenotypic characteristics, based on advanced multiparametric MR neuroimaging data and texture features, utilizing state-of-the-art radiomic analysis methods, can significantly contribute to the pre-treatment glioma grade differentiation.

AB - AIMS AND OBJECTIVES: To investigate the value of advanced multiparametric MR imaging biomarker analysis based on radiomic features and machine learning classification, in the non-invasive evaluation of tumor heterogeneity towards the differentiation of Low Grade vs. High Grade Gliomas.METHODS AND MATERIALS: Forty histologically confirmed glioma patients (20 LGG and 20 HGG) who underwent a standard 3T-MRI tumor protocol with conventional (T1 pre/post-contrast, T2-FSE, T2-FLAIR) and advanced techniques (Diffusion Tensor and Perfusion Imaging, 1H-MR Spectroscopy), were included. A semi-automated segmentation technique, based on T1W-C and DTI, was used for tumor core delineation in all available parametric maps. 3D Texture analysis considered 12 Histogram, 11 Co-Occurrence Matrix (GLCM) and 5 Run Length Matrix (GLRLM) features, derived from p, q, MD, FA, T1W-C, T2W-FSE, T2W-FLAIR and raw DSCE data. Along with 1H-MRS metabolic ratios and mean rCBV values, a total of 581 attributes for each subject were obtained. A Support Vector Machine - Recursive Feature Elimination (SVM-RFE) algorithm and SVM classifier were utilized for feature selection and classification, respectively.RESULTS: Three different SVM classifiers were evaluated with consecutively SVM-RFE feature subsets. Linear SMO classifier demonstrated the highest performance for determining the optimal feature subset. Finally, 21 SVM-RFE top-ranked features were adopted, for training and testing the SMO classifier with leave-one-out cross-validation, achieving 95.5% Accuracy, 95% Sensitivity, 96% Specificity and 95.5% Area Under ROC Curve.CONCLUSION: Results demonstrate that quantitative analysis of phenotypic characteristics, based on advanced multiparametric MR neuroimaging data and texture features, utilizing state-of-the-art radiomic analysis methods, can significantly contribute to the pre-treatment glioma grade differentiation.

KW - Glioma grading

KW - Machine learning

KW - Multiparametric MRI

KW - Radiomics

UR - http://www.scopus.com/inward/record.url?scp=85063205125&partnerID=8YFLogxK

U2 - 10.1016/j.ejmp.2019.03.014

DO - 10.1016/j.ejmp.2019.03.014

M3 - Article

C2 - 30910431

SN - 1120-1797

VL - 60

SP - 188

EP - 198

JO - Physica Medica

JF - Physica Medica

ER -

Imaging biomarker analysis of advanced multiparametric MRI for glioma grading

Abstract

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this