4 Way Classification of Brain Tumor Using Shallow Convolution Neural Network.

EasyChair Preprint 7084

Abstract

The range of Central Nervous System (CNS) tumors in India is from 5 to 10 per 100000 of the population on average that is increasing as the time passes. The medical services in India are still out of reach for many people. However, the author has the view that the advancement in the field of deep learning, machine learning and artificial intelligence can be utilised to solve this issue. In this paper, we are trying to identify three types of brain tumor namely maningioma, glioma, and pituitary while separating the no tumor category. Here we are using MRI scan images. These images are of jpeg form, pre-processed taken from kaggle database. The accuracy of our model in identifying the above mentioned four classes is 90.16%. Similarly, the precision for Glioma, meningioma, pituitary and no tumor are 81.48%, 78.94%, 94.73% and 97.36 % respectively.

Keyphrases: Artificial Intelligence, Brain Tumor, CT(computed Tomography), MRI(magnetic resonance imaging), central nervous system, deep learning, machine learning

@booklet{EasyChair:7084,
  author    = {Abhishekh Sawner and Shailja Shukla and Vandana Roy and Sourav Shukla and Sarang Kapoor},
  title     = {4 Way Classification of Brain Tumor Using Shallow Convolution Neural Network.},
  howpublished = {EasyChair Preprint 7084},
  year      = {EasyChair, 2021}}