International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 09 Issue: 04 | Apr 2022
p-ISSN: 2395-0072
www.irjet.net
A Review Paper on Automated Brain Tumor Detection Febi B S1, Dr. Deepesh Edwin.2 PG Student, Dept. of Electronics & Communication Engineering, LBSITW, Kerala, India Assistant Professor, Dept. of Electronics & Communication Engineering, LBSITW, Kerala, India ---------------------------------------------------------------------***--------------------------------------------------------------------1
2
Abstract - One of the major leading death causes in the
radiation risk to the human body. As tens of thousands of patients suffer from brain tumors each year, the use of deep learning techniques for the automatic detection and classification of brain tumors has become an area of interest. The common problems in manual detection of the brain tumors are the significant time requirements and the possibilities of misclassification due to the complexity of the problem.
world is brain tumor. Early-stage brain tumor diagnosis has recently become a major topic of research. Many researchers have made their own efforts in diagnosing and predicting brain tumors, and each technique has a different accuracy rate, which varies for different situations and using datasets. Patient’s survival rate can be increased by detecting the tumor at its early stages. Medical evidence shows that man-assisted manual classification can lead to inaccurate prognosis and diagnosis. This is mainly due to the diversity and similarity of the tumor and normal tissues. Recently, deep learning techniques to improve the accuracy of brain tumor detection and classification from magnetic resonance imaging (MRI) have shown good results. The main purpose of this review is to highlight some of the previous studies of deep learning used to predict brain tumor.
Automatic brain tumor detection and classification methods are needed to reduce the diagnostic time and avoid human errors before making any decisions. This paper presents a deep learning (DL) strategy for automatic brain tumor diagnosis. Deep learning has been applied to many applications, such as pattern classification object detection, speech recognition and other decisions. Different learning strategies are suggested to detect brain tumors.
Key Words: Brain Tumor, MRI Images, deep learning
2. REVIEW ON DIFFERENT PAPERS
1. INTRODUCTION
Yakub bhanothu et.al [4] this paper explains the automatic brain tumor detection and classification of MR images using the deep learning algorithm. The rapid R-CNN algorithm was used to locate tumor areas and classify them into three types: glioma, meningioma, and pituitary tumor. VGG-16 used a CNN architecture as the base layer for specific networks. Achieved improved mAP to detect brain tumors.
Medical problems related to the brain are the most critical and complex to diagnose and operate on. In India, statistics reveal 40,000 to 50,000 new brain tumor cases each year of which 20% are brain tumor cases in children. Abnormal cell growth in the human brain is called a brain tumor. There are many types of brain tumors. Some brain tumors are cancerous (malignant) and some are noncancerous (benign). A tumor that develops in the brain or spinal cord is called a glioma and a tumor that develops from the meninges is called a meningioma. Abnormal cell growth in the pituitary tumor.
Ekam singh Chahal et.al [1] brain tumor detection using deep learning model combines two-pathway and cascade architectures to analyze and implement brain segmentation. The input cascade performs better than the MFC cascade. Mohamed R Shoaib et.al [6] this paper describes a comparative study of four CNN based models for brain tumor detection. The findings are based on a four-class brain tumor MR image database. The CNN model based on transfer learning gives better results compared to the more advanced methods. The inception ResNet v2 model showed effective results with an accuracy of 86.80%.
Early detection of a brain tumor is essential for effective. For medical image diagnosis, image can be obtained from a variety of imaging techniques, including positron emission tomography (PET), magnetic resonance imaging (MRI), and computed tomography (CT). MRI was found to be the best diagnosis for brain tumors. Unlike CT scans, MRI does not use radiation of X-rays which can cause damage. MRI images have a high resolution and are very detailed so they can detect even small things.
Hassan ali khan et.al [5] brain tumor detection using three deep learning models mainly VGG-16, ResNet-50 and inception-V3 model. The VGG-16 model showed effective results with an accuracy of 96%.
The tumor may appear as a white area or as a bright white pattern. Even though, these cells have other parts of the brain that are similar in nature, which can lead to misdiagnosis. Now MR images are very useful in a medical image processing. MRI is harmless because it is based on magnetic field and radio waves and does not pose any
© 2022, IRJET
|
Impact Factor value: 7.529
Neba Sharma et.al [3] the VGG-16 pre-trained CNN model used the transfer learning approach with data enhancement to classify normal and abnormal brain MRI images and achieved 90% accuracy.
|
ISO 9001:2008 Certified Journal
|
Page 3908