Computer-Aided Grading of Gliomas Combining Automatic Segmentation and Radiomics

Wei Chen,Xu Qiao,Suting Peng,Jiawei Sun,Boqiang Liu

doi:10.1155/2018/2512037

Abstract

Gliomas are the most common primary brain tumors, and the objective grading is of great importance for treatment. This paper presents an automatic computer-aided diagnosis of gliomas that combines automatic segmentation and radiomics, which can improve the diagnostic ability. The MRI data containing 220 high-grade gliomas and 54 low-grade gliomas are used to evaluate our system. A multiscale 3D convolutional neural network is trained to segment whole tumor regions. A wide range of radiomic features including first-order features, shape features, and texture features is extracted. By using support vector machines with recursive feature elimination for feature selection, a CAD system that has an extreme gradient boosting classifier with a 5-fold cross-validation is constructed for the grading of gliomas. Our CAD system is highly effective for the grading of gliomas with an accuracy of 91.27%, a weighted macroprecision of 91.27%, a weighted macrorecall of 91.27%, and a weighted macro-F1 score of 90.64%. This demonstrates that the proposed CAD system can assist radiologists for high accurate grading of gliomas and has the potential for clinical applications.

Highlights

Gliomas are the most common primary brain tumors, characterized by the uncontrolled proliferation of abnormal brain cells
The methods involved in the computer-aided detection diagnosis (CAD) system are mainly introduced, which are organized as 5 parts: data preprocessing, automated tumor segmentation, radiomic features extraction, feature selection, and data classification
Consistent with the features described by Imaging Biomarker Standardization Initiative (IBSI) [17], a wide range of radiomic features including first-order features, shape features, and texture features was extracted from the segmented brain tumor regions

Summary

Introduction

Gliomas are the most common primary brain tumors, characterized by the uncontrolled proliferation of abnormal brain cells. There are several common MRI modalities including T1weighted (T1), T2-weighted (T2), gadolinium enhanced T1weighted (T1c), and Fluid-Attenuated Inversion Recovery (FLAIR), which generate a large number of medical images This has become a huge burden for radiologists, resulting in inaccurate detection or misinterpretation. The CAD system based on the use of automatic segmentation and ensemble classification techniques can accurately classify brain tumor as benign or malignant [8]. All of these studies have shown that CAD can help to improve the accuracy and efficiency of the diagnostic process and reduce the burden of work. The system is automated and general and, if a significant number of subjects exist, has the potential for clinical applications

Method

Result

Findings

Conclusion