MRI-Based Brain Tumor Classification Using Ensemble of Deep Features and Machine Learning Classifiers.

Jaeyong Kang,Zahid Ullah,Jeonghwan Gwak

doi:10.3390/s21062222

Jaeyong Kang, Zahid Ullah + Show 1 more

Open Access

https://doi.org/10.3390/s21062222

Copy DOI

Abstract

Brain tumor classification plays an important role in clinical diagnosis and effective treatment. In this work, we propose a method for brain tumor classification using an ensemble of deep features and machine learning classifiers. In our proposed framework, we adopt the concept of transfer learning and uses several pre-trained deep convolutional neural networks to extract deep features from brain magnetic resonance (MR) images. The extracted deep features are then evaluated by several machine learning classifiers. The top three deep features which perform well on several machine learning classifiers are selected and concatenated as an ensemble of deep features which is then fed into several machine learning classifiers to predict the final output. To evaluate the different kinds of pre-trained models as a deep feature extractor, machine learning classifiers, and the effectiveness of an ensemble of deep feature for brain tumor classification, we use three different brain magnetic resonance imaging (MRI) datasets that are openly accessible from the web. Experimental results demonstrate that an ensemble of deep features can help improving performance significantly, and in most cases, support vector machine (SVM) with radial basis function (RBF) kernel outperforms other machine learning classifiers, especially for large datasets.

Highlights

In the human body, the brain is an enormous and complex organ that controls the whole nervous system, and it contains around 100-billion nerve cells [1]
We evaluate each deep feature extracted from 13 different pre-trained convolutional neural networks (CNNs) using 9 different machine learning (ML) classifiers (FC, Gaussian NB, AdaBoost, K-Nearest Neighbors (k-NN), Random forest (RF), support vector machine (SVM)-linear, SVM-sigmoid, SVM-radial basis function (RBF), and Extreme Learning Machine (ELM)) described in Section 3.3 and choose the top three deep features based on the average accuracy of 9 different ML classifiers for each of our 3 different magnetic resonance imaging (MRI) datasets
The second experiment is designed to show the effectiveness of the ensemble of top 2 or 3 deep features selected by the results from the first experiment with several different ML classifiers

Summary

Introduction

The brain is an enormous and complex organ that controls the whole nervous system, and it contains around 100-billion nerve cells [1]. This essential organ is originated in the center of the nervous system. Any kind of abnormality that exists in the brain may put human health in danger. Among such abnormalities, brain tumors are the most severe ones. The primary tumors present in the brain tissue, while the secondary tumors expand from other parts of the human body to the brain tissue through the bloodstream [2]. The most common brain tumor in humans is glioma [4]

Methods

Results

Discussion

Conclusion