Brain tumor segmentation in multi‐spectral MRI using convolutional neural networks (CNN)

Abstract

A tumor could be found in any area of the brain and could be of any size, shape, and contrast. There may exist multiple tumors of different types in a human brain at the same time. Accurate tumor area segmentation is considered primary step for treatment of brain tumors. Deep Learning is a set of promising techniques that could provide better results as compared to nondeep learning techniques for segmenting timorous part inside a brain. This article presents a deep convolutional neural network (CNN) to segment brain tumors in MRIs. The proposed network uses BRATS segmentation challenge dataset which is composed of images obtained through four different modalities. Accordingly, we present an extended version of existing network to solve segmentation problem. The network architecture consists of multiple neural network layers connected in sequential order with the feeding of Convolutional feature maps at the peer level. Experimental results on BRATS 2015 benchmark data thus show the usability of the proposed approach and its superiority over the other approaches in this area of research.

Multiclassifier for severity‐level categorization of glioma tumors using multimodal magnetic resonance imaging brain images

“…However, the method required to bias the ground truth labels systematically. Moreover, segmentation based on the DCNN was developed by Iqbal, 34 which was reported to have large training data and high computational cost. Gupta et al 27 used an adaptive thresholding for identifying the gliomas, which offered poor performance in noisy situations and required prior knowledge regarding segmentation.…”

“…One of the emerging methods for multilevel classification is DCNN, 34 which is an efficacious method to model the classes distinguished with nonlinear boundaries when compared with other classifiers such as support vector machine (SVM), feed forward neural network (FFNN), and fuzzy classifier. The segmentation using random forest needed a priori knowledge and took huge computational effort.…”

Fuzzy ARTMAP Approach for Arabic Writer Identification using Novel Features Fusion

“…Further segmentation of the images is desired priror to features extraction (Rahim et al, 2017a;Iqbal et al, 2018;2017). In this study, since only author identification is required, actual meanings of the text and knowledge of characters involved remain of lesser importance.…”

An ensemble classification of exudates in color fundus images using an evolutionary algorithm based optimal features selection

“…The main contribution achieved in this article could be summarized as: Salient shape and texture based features are extracted from optic fundus images based on accurate segmentation techniques to detect EXs. Optimization of features is performed through the use of evolutionary Genetic Algorithm (GA), which results in the optimal selection of best features from existing features. The optimization result is an upgradation in terms of computational performance (Rad, Rahim, Rehman, & Saba, ). Three well‐known classifiers [Naïve Bayesian (NB), SVM, and Artificial Neural Network (ANN)] are trained to classify EXs in optic fundus images. Similarly, an ensemble based classifier is used to select the best classifier for EXs classification in fundus images through majority voting scheme (Iqbal, Khan, Saba, & Rehman, ; Iqbal, Ghani, Saba, & Rehman, ). …”