Detection of Abnormality in Brain Images using Discrete Ridgelet Transform
K. Susmitha Devi1, G. Prathibha.2
1K. Susmitha Devi, PG Student, Dept. of Engineering and Communication Engineering, Acharya Nagarjuna University, Guntur, Andhra Pradesh.
2Dr. G. Prathibha, Asst. Prof… Dept. of Engineering and Communication Engineering, Acharya Nagarjuna University, Guntur, Andhra Pradesh.
Manuscript received on 20 April 2019 | Revised Manuscript received on 26 May 2019 | Manuscript published on 30 May 2019 | PP: 1852-1857 | Volume-8 Issue-1, May 2019 | Retrieval Number: A1172058119/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: TClassification plays a key role for differentiate the brain images as normal and abnormal. Abnormal images mean the brain images contains hematomas (bleeding of blood outside of the blood vessels), tumors (growth of unwanted tissues in blood vessels) etc. Brain Atlas Database is used for proposed work this database comprises normal and abnormal images with different plane like axial, sagittal and coronal with T1&T2 weighted images The present study is used to classify the MRI brain images with kernel SVM incorporated with 3 stages: Preprocessing, Feature extraction, training and classification. In preprocessing grayscale conversion and median filtering is done. Ridgelet Transform is proposed specifically for directional features with line singularities and con-entropy features are extracted by the GLCM. These feature are combined as a feature set and fed input to the kernel support vector machine for training and classification it classifies the brain images with different levels of accuracies. Experimental results show the kernel SVM yields better accuracy compared with existing classifiers.
Index Terms: Brain MRI, Ridgelet Transform, GLCM, Classification, Kernel Support Vector Machine.
Scope of the Article: Classification