References | Purpose | Model | Limitations/future works |
---|---|---|---|
Badža et al. [5] | To classify different types of brain tumors using a convolutional neural network | CNN | Examining the execution of the designed neural network in the mentioned study, as well as enhanced ones in different medical images |
Gumaei et al. [23] | To classify brain tumors using a hybrid feature extraction method | RELM | Lack of comparison of the technique used in this study and other machine learning methods |
Rehman et al. [22] | Proposing three architectures of convolutional neural networks (alexnet,Googlenet, and vggnet) to classify brain tumors | Convolutional neural networks (AlexNet, GoogLeNet, and VGGNet) | Explore other essential deep neural network’s architectures for brain tumor classification with less time complexity |
Mittal et al. [29] | Using segmentation method to diagnose brain tumors using deep learning-based methods | Combination of SWT and GCNN | Other databases like PASCAL, Berkeley or BRATS can be used |
It is recommended to use a variety of diverse classifiers to increase the accuracy of the classifier | |||
Phaye et al. [24] | Provide an approach to improve outputs using a network with dense layers | Dense capsule networks (DCNet) and diverse capsule networks (DCNet++) | Computational complexity must be reduced to enhance classifier execution |
Pashaei et al. [27] | Developing an algorithm for extracting and classifying features with the CNN and KELM | KELM | Not mentioned |
Abiwinanda et al. [28] | Use the convolutional neural network to segment and classify brain tumors automatically | CNN | Pay attention to the color balancing step to improve the classifier's accuracy |
Abd-Ellah et al. [15] | Brain tumor detection with a two-step automatic detection system | Preprocessing, feature extraction using CNN and classification with error-correcting output codes support vector machine (ECOC-SVM) | Not mentioned |