Classification of Magnetic Resonance Images Using Eight Directions Gray Level Co-Occurrence Matrix (8dglcm) Based Feature Extraction
P. Santhi1, G. Mahalakshmi2

1P. Santhi, Department of CSE, M. Kumarasamy College of Engineering, Karur (Tamil Nadu), India.
2G. Mahalakshmi, Department of CSE, M. Kumarasamy College of Engineering, Karur (Tamil Nadu), India.

Manuscript received on 18 April 2019 | Revised Manuscript received on 25 April 2019 | Manuscript published on 30 April 2019 | PP: 839-845 | Volume-8 Issue-4, April 2019 | Retrieval Number: D6397048419/19©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 (

Abstract: Classification of MRI images is very difficult due to the variance and various complexities of tumor cells. The proposed classification system is designed for differentiating the brain MRI images into three classes such as Malignant, Benign and Normal. The proposed probability based Support Vector Machine (SVM) includes characteristic Extraction, Best Feature Subset Selection and Classification. In Feature Extraction, most of researchers are using GLCM method for extracting the texture features from an image. Main limitation of GLCM is that, it is computationally very intensive and many of the calculations are done using unnecessary zero frequencies. To avoid the limitations of GLCM, this paper introduces the 8DGLCM for feature extraction. Performance of classifiers is reduced if many features are considered during object identification. Feature Selection method is used to deal with the issues in feature dimensionality by way of selecting the best features subset. Here, Ranking based Particle Swarm Optimization (PSO) is concentrates to choose the best feature subset from an extracted feature. Finally, the MRI images are classified using the probability based SVM classifier. The performance of this method is evaluated based on 7 MRI image sets. An expert radiologist observation is used as reference to evaluate the performance of this system. Final result shows the performance of proposed system is 95.65%.
Keywords: Feature Extraction, Selection, Classification, Swarm Optimization, Co-Occurrence Matrix.

Scope of the Article: Classification