Brain Tumor Detection Using Convolution Neura...

Brain Tumor Detection Using Convolution Neural Network

Author Name : Anushka Pathak

ABSTRACT

Human-assisted detection can result in inaccurate prediction and diagnosis, making brain tumor detection one of the most important and difficult tasks in the field of medical image processing. This is because brain tumors have a range of appearances, and since tumor and normal tissues are similar, it is hard to tell which regions belong to a tumor. I suggest utilizing a CNN for brain tumor detection in 2D MRI images due to its effectiveness in image classification, which could improve diagnosis and treatment outcomes. The experiment was conducted on a dataset that included tumours of various sizes and locations. Since CNN performs better than traditional methods, Since it performs better than conventional methods, CNN was introduced using Keras and Tensor Flow. Keras and Tensor Flow were preferred over traditional methods for implementing CNN due to their capability to achieve better performance. In proposed work, CNN achieved an accuracy of 84.4%, which is very compelling considering use of a custom model and not a pre-trained one. The main aim of this paper is to improve brain tumor detection using Convolutional Neural Networks.

Keywords: Data Augmentation; Brain Tumor; Deep Learning; Convolution Neural Network; Relu Activation Function