Brain Tumor Detection Using Hybrid Model

Brain Tumor Detection Using Hybrid Model

Anik Datta¹, Ashutosh Luthra², Arpita Kanaujia³, Sanjeev Kumar⁴

¹Department of Computer Science and Engineering, ABES Institute of Technology, Ghaziabad,     Affiliated to AKTU Lucknow India

² Department of Computer Science and Engineering, ABES Institute of Technology, Ghaziabad,     Affiliated to AKTU Lucknow India

³ Department of Computer Science and Engineering, ABES Institute of Technology, Ghaziabad,     Affiliated to AKTU Lucknow India

⁴ Department of Computer Science and Engineering, ABES Institute of Technology, Ghaziabad,    Affiliated to AKTU Lucknow India

Abstract- Computer-assisted diagnosis (CAD) has become more significant in medical applications to assist doctors in making appropriate judgments. One such application is the identification of brain tumours using MRI images of the brain. Several machine learning techniques have been developed to categorise brain MRI with varying degrees of accuracy. A dense net neural network model is utilised in this study to categorise a brain MRI picture into two classes: tumour and no tumour [1]. Identifying MRI brain tumours is among the most essential tasks in medical image analysis. A brain tumour is an uncontrollably growing and multiplying cluster of cells. It is created by unfavourable cells that are typically present in various parts of the brain, including as lymphatic tissue, glial cells, neurons, blood vessels, and the skull, or it is disseminated from tumours that are mostly present in other systems. To increase the models accuracy, the original dense net model is expanded with certain connection alterations in this study. When compared to previous deep learning models, the updated dense net model can acquire detailed information from an MRI picture and categorise it with more accuracy.

Keywords: Tumor, Support Vector Machine (SVM), Convolutional Neural Network (CNN), Hybrid-Model.