Classification and Extraction of Brain Tumor using Hybrid Algorithm

Abstract

Rapid growth of unwanted cells or abnormal accumulation of tissues that leads in the formation of tumor in brain is called Brain tumor. Tumor is always considered to be a dreadful disease, if not detected at the initial stage. As this has been a rapid growing threat among human race. Detection and extraction of such vulnerable dangerous diseases has become the most demanding one. In spite of existing techniques of detection, the accuracy has always been a challenging task. Misinterpretation is occurred during the evaluation of data sets of images, in detection of tumor which leads to a contradictory validation. Extraction of the tumor region from large data sets is has always been a grim. The extant innovations are Non- negative factorization (NMF) Principle Component Analysis (PCA), Convolutional Neural Network (CNN), K-means clustering algorithm, one-way analysis and variance (ANOVA) etc. In this research paper a proposed solution for the classification and extraction has been resolved by using hybrid algorithm. The hybrid consists of two algorithms further classified into two modules which consist of Convolutional Neural Network (CNN) and Deep Neural Network (DNN) respectively.

Keywords: classification, CNN, DNN, extraction, hybrid, tumor.