Texture analysis-and support vector machine-assisted diffusional kurtosis imaging may allow in vivo gliomas grading and IDH-mutation status prediction

a preliminary study

Sotirios Bisdas, Haocheng Shen, Stef Thust, Vasileios Katsaros, George Stranjalis, Christos Boskos, Sebastian Brandner, Jianguo Zhang

Research output: Contribution to journalArticle

7 Citations (Scopus)
38 Downloads (Pure)

Abstract

We sought to investigate, whether texture analysis of diffusional kurtosis imaging (DKI) enhanced by support vector machine (SVM) analysis may provide biomarkers for gliomas staging and detection of the IDH mutation. First-order statistics and texture feature extraction were performed in 37 patients on both conventional (FLAIR) and mean diffusional kurtosis (MDK) images and recursive feature elimination (RFE) methodology based on SVM was employed to select the most discriminative diagnostic biomarkers. The first-order statistics demonstrated significantly lower MDK values in the IDH-mutant tumors. This resulted in 81.1% accuracy (sensitivity = 0.96, specificity = 0.45, AUC 0.59) for IDH mutation diagnosis. There were non-significant differences in average MDK and skewness among the different tumour grades. When texture analysis and SVM were utilized, the grading accuracy achieved by DKI biomarkers was 78.1% (sensitivity 0.77, specificity 0.79, AUC 0.79); the prediction accuracy for IDH mutation reached 83.8% (sensitivity 0.96, specificity 0.55, AUC 0.87). For the IDH mutation task, DKI outperformed significantly the FLAIR imaging. When using selected biomarkers after RFE, the prediction accuracy achieved 83.8% (sensitivity 0.92, specificity 0.64, AUC 0.88). These findings demonstrate the superiority of DKI enhanced by texture analysis and SVM, compared to conventional imaging, for gliomas staging and prediction of IDH mutational status.

Original languageEnglish
Article number6108
Pages (from-to)1-9
Number of pages9
JournalScientific Reports
Volume8
Issue number1
DOIs
Publication statusPublished - 17 Apr 2018

Fingerprint

Glioma
Area Under Curve
Biomarkers
Sensitivity and Specificity
Mutation
Neoplasms
Support Vector Machine

Keywords

  • Cancer imaging
  • CNS cancer

Cite this

Bisdas, Sotirios ; Shen, Haocheng ; Thust, Stef ; Katsaros, Vasileios ; Stranjalis, George ; Boskos, Christos ; Brandner, Sebastian ; Zhang, Jianguo. / Texture analysis-and support vector machine-assisted diffusional kurtosis imaging may allow in vivo gliomas grading and IDH-mutation status prediction : a preliminary study. In: Scientific Reports. 2018 ; Vol. 8, No. 1. pp. 1-9.
@article{f230336fbc034caaa387772f94d7d488,
title = "Texture analysis-and support vector machine-assisted diffusional kurtosis imaging may allow in vivo gliomas grading and IDH-mutation status prediction: a preliminary study",
abstract = "We sought to investigate, whether texture analysis of diffusional kurtosis imaging (DKI) enhanced by support vector machine (SVM) analysis may provide biomarkers for gliomas staging and detection of the IDH mutation. First-order statistics and texture feature extraction were performed in 37 patients on both conventional (FLAIR) and mean diffusional kurtosis (MDK) images and recursive feature elimination (RFE) methodology based on SVM was employed to select the most discriminative diagnostic biomarkers. The first-order statistics demonstrated significantly lower MDK values in the IDH-mutant tumors. This resulted in 81.1{\%} accuracy (sensitivity = 0.96, specificity = 0.45, AUC 0.59) for IDH mutation diagnosis. There were non-significant differences in average MDK and skewness among the different tumour grades. When texture analysis and SVM were utilized, the grading accuracy achieved by DKI biomarkers was 78.1{\%} (sensitivity 0.77, specificity 0.79, AUC 0.79); the prediction accuracy for IDH mutation reached 83.8{\%} (sensitivity 0.96, specificity 0.55, AUC 0.87). For the IDH mutation task, DKI outperformed significantly the FLAIR imaging. When using selected biomarkers after RFE, the prediction accuracy achieved 83.8{\%} (sensitivity 0.92, specificity 0.64, AUC 0.88). These findings demonstrate the superiority of DKI enhanced by texture analysis and SVM, compared to conventional imaging, for gliomas staging and prediction of IDH mutational status.",
keywords = "Cancer imaging, CNS cancer",
author = "Sotirios Bisdas and Haocheng Shen and Stef Thust and Vasileios Katsaros and George Stranjalis and Christos Boskos and Sebastian Brandner and Jianguo Zhang",
note = "no funding info",
year = "2018",
month = "4",
day = "17",
doi = "10.1038/s41598-018-24438-4",
language = "English",
volume = "8",
pages = "1--9",
journal = "Scientific Reports",
issn = "2045-2322",
publisher = "Nature Publishing Group",
number = "1",

}

Texture analysis-and support vector machine-assisted diffusional kurtosis imaging may allow in vivo gliomas grading and IDH-mutation status prediction : a preliminary study. / Bisdas, Sotirios ; Shen, Haocheng; Thust, Stef ; Katsaros, Vasileios ; Stranjalis, George ; Boskos, Christos ; Brandner, Sebastian ; Zhang, Jianguo.

In: Scientific Reports, Vol. 8, No. 1, 6108, 17.04.2018, p. 1-9.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Texture analysis-and support vector machine-assisted diffusional kurtosis imaging may allow in vivo gliomas grading and IDH-mutation status prediction

T2 - a preliminary study

AU - Bisdas, Sotirios

AU - Shen, Haocheng

AU - Thust, Stef

AU - Katsaros, Vasileios

AU - Stranjalis, George

AU - Boskos, Christos

AU - Brandner, Sebastian

AU - Zhang, Jianguo

N1 - no funding info

PY - 2018/4/17

Y1 - 2018/4/17

N2 - We sought to investigate, whether texture analysis of diffusional kurtosis imaging (DKI) enhanced by support vector machine (SVM) analysis may provide biomarkers for gliomas staging and detection of the IDH mutation. First-order statistics and texture feature extraction were performed in 37 patients on both conventional (FLAIR) and mean diffusional kurtosis (MDK) images and recursive feature elimination (RFE) methodology based on SVM was employed to select the most discriminative diagnostic biomarkers. The first-order statistics demonstrated significantly lower MDK values in the IDH-mutant tumors. This resulted in 81.1% accuracy (sensitivity = 0.96, specificity = 0.45, AUC 0.59) for IDH mutation diagnosis. There were non-significant differences in average MDK and skewness among the different tumour grades. When texture analysis and SVM were utilized, the grading accuracy achieved by DKI biomarkers was 78.1% (sensitivity 0.77, specificity 0.79, AUC 0.79); the prediction accuracy for IDH mutation reached 83.8% (sensitivity 0.96, specificity 0.55, AUC 0.87). For the IDH mutation task, DKI outperformed significantly the FLAIR imaging. When using selected biomarkers after RFE, the prediction accuracy achieved 83.8% (sensitivity 0.92, specificity 0.64, AUC 0.88). These findings demonstrate the superiority of DKI enhanced by texture analysis and SVM, compared to conventional imaging, for gliomas staging and prediction of IDH mutational status.

AB - We sought to investigate, whether texture analysis of diffusional kurtosis imaging (DKI) enhanced by support vector machine (SVM) analysis may provide biomarkers for gliomas staging and detection of the IDH mutation. First-order statistics and texture feature extraction were performed in 37 patients on both conventional (FLAIR) and mean diffusional kurtosis (MDK) images and recursive feature elimination (RFE) methodology based on SVM was employed to select the most discriminative diagnostic biomarkers. The first-order statistics demonstrated significantly lower MDK values in the IDH-mutant tumors. This resulted in 81.1% accuracy (sensitivity = 0.96, specificity = 0.45, AUC 0.59) for IDH mutation diagnosis. There were non-significant differences in average MDK and skewness among the different tumour grades. When texture analysis and SVM were utilized, the grading accuracy achieved by DKI biomarkers was 78.1% (sensitivity 0.77, specificity 0.79, AUC 0.79); the prediction accuracy for IDH mutation reached 83.8% (sensitivity 0.96, specificity 0.55, AUC 0.87). For the IDH mutation task, DKI outperformed significantly the FLAIR imaging. When using selected biomarkers after RFE, the prediction accuracy achieved 83.8% (sensitivity 0.92, specificity 0.64, AUC 0.88). These findings demonstrate the superiority of DKI enhanced by texture analysis and SVM, compared to conventional imaging, for gliomas staging and prediction of IDH mutational status.

KW - Cancer imaging

KW - CNS cancer

UR - http://www.scopus.com/inward/record.url?scp=85045762290&partnerID=8YFLogxK

U2 - 10.1038/s41598-018-24438-4

DO - 10.1038/s41598-018-24438-4

M3 - Article

VL - 8

SP - 1

EP - 9

JO - Scientific Reports

JF - Scientific Reports

SN - 2045-2322

IS - 1

M1 - 6108

ER -