An Efficient Method for Diagnosing Brain Tumors Based on MRI Images Using Deep Convolutional Neural Networks

Table 1

Summary of studies on brain tumor classification.

Author	Classification method	Objective	Dataset	Feature extraction method	Accuracy
Khawaldeh et al. [8]	CNN	Classification of brain MRI into normal and abnormal	587 MR images	CNN	91.16%
Paul et al. [9]	Fully connected and CNN	Brain tumor classification of MR brain image	3064 MR images	CNN	91.43%
Varuna Shree and Kumar [10]	Probabilistic neural network (PNN)	Classification of brain MRI into normal and abnormal	650 MR images	Gray level cooccurrence matrix (GLCM)	95%
Hemanth et al. [11]	CNN	Classification into normal and abnormal	220 MR images	CNN	94.5%
Deepak and Ameer [12]	Deep transfer learning	Classification of glioma, meningioma, and pituitary tumors	3064 MR images	Google Net	98%
Das et al. [13]	CNN	Brain tumor classification	3064 MR images	CNN	94.39%
Ullah et al. [14]	Feedforward neural network	Classification of brain MRI into normal and abnormal	71 MR images	DWT	95.8%
Çinar and Yildirim [15]	CNN models	Brain tumor detection and classification	253 MR images	CNN	97.2%
Siddiaue et al. [16]	Proposed DCNN model	Brain tumor classification	253 MR images	CNN	96%
Saxena et al. [17]	CNN networks with transfer learning	Binary classification of brain tumor normal and abnormal	253 MR images	CNN	95%
Díaz-Pernas et al. [18]	Multipathway CNN	Brain tumor classification	3064 MR images	CNN	97.3%
Abd El kader et al. [19]	Proposed differential deep-CNN	Brain tumor classification	25000 MR images	CNN	99.25%
Tazin et al. [20]	CNN architectures	Brain tumor classification	2513 MR images	CNN	Up to 92%