Fusion Based Brain Tumor Detection Using Machine Learning

Kamali R, Dommaraju Divija, Eduru Madhavi Reddy, Dr. Thilagavathy A

Abstract: t: The growth abnormal cells in the brain is the main cause for Brain Tumor. There are various types of brain tumor that exist. The brain tumor is mainly classified as Benign and Malignant. The rate at which the brain tumor grows and its location may vary greatly which also determines how it could affect the function of the nervous system of the effected person. In this paper we use fusion of medical images (MRI image) and machine learning (SVM) for the diagnosis and classification of the type of brain tumor. Fusion of image is used in the reduction of both uncertainty and redundancy while extracting every possible useful information from that particular image. Support Vector Machine is a supervised machine learning algorithm which is used to provide more efficiency in the field of classification. As a result fused image contains more information than that of the source images. Fused images extract the features that involve texture and wavelet. The SVM classifier mainly classifies the tumor based on the extracted, trained and tested features. The experimental result proves that better performance would be achieved by using fusion based brain tumor detection using machine learning which has been proposed in this paper. ?

Keywords: : Machine Learning, Feature Extraction, SVM, Image Processing, MRI, Segmenting