Radiomics MRI phenotyping with machine learning to predict the grade of lower-grade gliomas: A study focused on nonenhancing tumors

Yae Won Park, Yoon Seong Choi, Sung Soo Ahn, Jong Hee Chang, Se Hoon Kim, Seung Koo Lee

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Objective: To assess whether radiomics features derived from multiparametric MRI can predict the tumor grade of lower-grade gliomas (LGGs; World Health Organization grade II and grade III) and the nonenhancing LGG subgroup. Materials and Methods: Two-hundred four patients with LGGs from our institutional cohort were allocated to training (n = 136) and test (n = 68) sets. Postcontrast T1-weighted images, T2-weighted images, and fluid-attenuated inversion recovery images were analyzed to extract 250 radiomics features. Various machine learning classifiers were trained using the radiomics features to predict the glioma grade. The trained classifiers were internally validated on the institutional test set and externally validated on a separate cohort (n = 99) from The Cancer Genome Atlas (TCGA). Classifier performance was assessed by determining the area under the curve (AUC) from receiver operating characteristic curve analysis. An identical process was performed in the nonenhancing LGG subgroup (institutional training set, n = 73; institutional test set, n = 37; and TCGA cohort, n = 37) to predict the glioma grade. Results: The performance of the best classifier was good in the internal validation set (AUC, 0.85) and fair in the external validation set (AUC, 0.72) to predict the LGG grade. For the nonenhancing LGG subgroup, the performance of the best classifier was good in the internal validation set (AUC, 0.82), but poor in the external validation set (AUC, 0.68). Conclusion: Radiomics feature-based classifiers may be useful to predict LGG grades. However, radiomics classifiers may have a limited value when applied to the nonenhancing LGG subgroup in a TCGA cohort.

Original languageEnglish
Pages (from-to)1381-1389
Number of pages9
JournalKorean journal of radiology
Volume20
Issue number9
DOIs
Publication statusPublished - 2019 Sep

Fingerprint

Glioma
Area Under Curve
Atlases
Genome
Neoplasms
ROC Curve
Machine Learning

All Science Journal Classification (ASJC) codes

  • Radiology Nuclear Medicine and imaging

Cite this

Park, Yae Won ; Choi, Yoon Seong ; Ahn, Sung Soo ; Chang, Jong Hee ; Kim, Se Hoon ; Lee, Seung Koo. / Radiomics MRI phenotyping with machine learning to predict the grade of lower-grade gliomas : A study focused on nonenhancing tumors. In: Korean journal of radiology. 2019 ; Vol. 20, No. 9. pp. 1381-1389.
@article{e358faae91dd43ca9df058c0e115f9e3,
title = "Radiomics MRI phenotyping with machine learning to predict the grade of lower-grade gliomas: A study focused on nonenhancing tumors",
abstract = "Objective: To assess whether radiomics features derived from multiparametric MRI can predict the tumor grade of lower-grade gliomas (LGGs; World Health Organization grade II and grade III) and the nonenhancing LGG subgroup. Materials and Methods: Two-hundred four patients with LGGs from our institutional cohort were allocated to training (n = 136) and test (n = 68) sets. Postcontrast T1-weighted images, T2-weighted images, and fluid-attenuated inversion recovery images were analyzed to extract 250 radiomics features. Various machine learning classifiers were trained using the radiomics features to predict the glioma grade. The trained classifiers were internally validated on the institutional test set and externally validated on a separate cohort (n = 99) from The Cancer Genome Atlas (TCGA). Classifier performance was assessed by determining the area under the curve (AUC) from receiver operating characteristic curve analysis. An identical process was performed in the nonenhancing LGG subgroup (institutional training set, n = 73; institutional test set, n = 37; and TCGA cohort, n = 37) to predict the glioma grade. Results: The performance of the best classifier was good in the internal validation set (AUC, 0.85) and fair in the external validation set (AUC, 0.72) to predict the LGG grade. For the nonenhancing LGG subgroup, the performance of the best classifier was good in the internal validation set (AUC, 0.82), but poor in the external validation set (AUC, 0.68). Conclusion: Radiomics feature-based classifiers may be useful to predict LGG grades. However, radiomics classifiers may have a limited value when applied to the nonenhancing LGG subgroup in a TCGA cohort.",
author = "Park, {Yae Won} and Choi, {Yoon Seong} and Ahn, {Sung Soo} and Chang, {Jong Hee} and Kim, {Se Hoon} and Lee, {Seung Koo}",
year = "2019",
month = "9",
doi = "10.3348/kjr.2018.0814",
language = "English",
volume = "20",
pages = "1381--1389",
journal = "Korean Journal of Radiology",
issn = "1229-6929",
publisher = "Korean Radiological Society",
number = "9",

}

Radiomics MRI phenotyping with machine learning to predict the grade of lower-grade gliomas : A study focused on nonenhancing tumors. / Park, Yae Won; Choi, Yoon Seong; Ahn, Sung Soo; Chang, Jong Hee; Kim, Se Hoon; Lee, Seung Koo.

In: Korean journal of radiology, Vol. 20, No. 9, 09.2019, p. 1381-1389.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Radiomics MRI phenotyping with machine learning to predict the grade of lower-grade gliomas

T2 - A study focused on nonenhancing tumors

AU - Park, Yae Won

AU - Choi, Yoon Seong

AU - Ahn, Sung Soo

AU - Chang, Jong Hee

AU - Kim, Se Hoon

AU - Lee, Seung Koo

PY - 2019/9

Y1 - 2019/9

N2 - Objective: To assess whether radiomics features derived from multiparametric MRI can predict the tumor grade of lower-grade gliomas (LGGs; World Health Organization grade II and grade III) and the nonenhancing LGG subgroup. Materials and Methods: Two-hundred four patients with LGGs from our institutional cohort were allocated to training (n = 136) and test (n = 68) sets. Postcontrast T1-weighted images, T2-weighted images, and fluid-attenuated inversion recovery images were analyzed to extract 250 radiomics features. Various machine learning classifiers were trained using the radiomics features to predict the glioma grade. The trained classifiers were internally validated on the institutional test set and externally validated on a separate cohort (n = 99) from The Cancer Genome Atlas (TCGA). Classifier performance was assessed by determining the area under the curve (AUC) from receiver operating characteristic curve analysis. An identical process was performed in the nonenhancing LGG subgroup (institutional training set, n = 73; institutional test set, n = 37; and TCGA cohort, n = 37) to predict the glioma grade. Results: The performance of the best classifier was good in the internal validation set (AUC, 0.85) and fair in the external validation set (AUC, 0.72) to predict the LGG grade. For the nonenhancing LGG subgroup, the performance of the best classifier was good in the internal validation set (AUC, 0.82), but poor in the external validation set (AUC, 0.68). Conclusion: Radiomics feature-based classifiers may be useful to predict LGG grades. However, radiomics classifiers may have a limited value when applied to the nonenhancing LGG subgroup in a TCGA cohort.

AB - Objective: To assess whether radiomics features derived from multiparametric MRI can predict the tumor grade of lower-grade gliomas (LGGs; World Health Organization grade II and grade III) and the nonenhancing LGG subgroup. Materials and Methods: Two-hundred four patients with LGGs from our institutional cohort were allocated to training (n = 136) and test (n = 68) sets. Postcontrast T1-weighted images, T2-weighted images, and fluid-attenuated inversion recovery images were analyzed to extract 250 radiomics features. Various machine learning classifiers were trained using the radiomics features to predict the glioma grade. The trained classifiers were internally validated on the institutional test set and externally validated on a separate cohort (n = 99) from The Cancer Genome Atlas (TCGA). Classifier performance was assessed by determining the area under the curve (AUC) from receiver operating characteristic curve analysis. An identical process was performed in the nonenhancing LGG subgroup (institutional training set, n = 73; institutional test set, n = 37; and TCGA cohort, n = 37) to predict the glioma grade. Results: The performance of the best classifier was good in the internal validation set (AUC, 0.85) and fair in the external validation set (AUC, 0.72) to predict the LGG grade. For the nonenhancing LGG subgroup, the performance of the best classifier was good in the internal validation set (AUC, 0.82), but poor in the external validation set (AUC, 0.68). Conclusion: Radiomics feature-based classifiers may be useful to predict LGG grades. However, radiomics classifiers may have a limited value when applied to the nonenhancing LGG subgroup in a TCGA cohort.

UR - http://www.scopus.com/inward/record.url?scp=85071636365&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85071636365&partnerID=8YFLogxK

U2 - 10.3348/kjr.2018.0814

DO - 10.3348/kjr.2018.0814

M3 - Article

C2 - 31464116

AN - SCOPUS:85071636365

VL - 20

SP - 1381

EP - 1389

JO - Korean Journal of Radiology

JF - Korean Journal of Radiology

SN - 1229-6929

IS - 9

ER -