ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

DEEP LEARNING-BASED BRAIN TUMOR DETECTION

Journal: International Journal of Advanced Research (Vol.10, No. 07)

Publication Date:

Authors : ; ;

Page : 552-562

Keywords : ;

Source : Downloadexternal Find it from : Google Scholarexternal

Abstract

A group of aberrant brain cells is known as a brain tumor. Brain tumors can be either malignant or benign. Whereas benign is not cancerous, malignant is. Both tumors are extremely dangerous because they spread quickly and affect many areas of the brain. About 166,039 people in the US had brain tumors or other cancers of the central nervous system in 2015.Even afterseveral researches,the cause of braintumorisunknown. We aimed to detect the brain tumor through image processing usingimages of MRI scans. Using Convolutional Neural Network(CNN) to perform imagebased classification. On receiving the image from the MRI scan as input, CNN will assess if the person has a brain tumor or not.It is observed on experimentation that the proposed approach needs an input image of size of240*240. Any image of another size cannot be used and the input image should only be a grayscale image. Color image cannot be used as an input image. The model which we proposed canbe further improved by using any size image as input. Moreover, the accuracy which we gotthrough this CNN model is 0.88. The model can be further improved by some changes so thataccuracy can be increased further.

Last modified: 2022-08-16 21:01:06