Experimental Approaches for Detection of Brain Tumor Grade Using Svm Classification
V Ramakrishna Sajja1, Gnaneswara Rao Nitta2
1V Ramakrishna Sajja, Department of Computer Science and Engineering, VFSTR Deemed to be University, Vadlamudi, Guntur (Andhra Pradesh), India.
2Gnaneswara Rao Nitta, Department of Computer Science and Engineering, VFSTR Deemed to be University, Vadlamudi, Guntur (Andhra Pradesh), India.
Manuscript received on 12 February 2019 | Revised Manuscript received on 02 March 2019 | Manuscript Published on 08 June 2019 | PP: 79-85 | Volume-7 Issue-5S4, February 2019 | Retrieval Number: E10160275S419/19©BEIESP
Open Access | Editorial and Publishing 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: Brain tumor detection is an urgent assignment for doctors. Cerebrum tumor grows quickly and average volume will be doubled in only twenty days. If it is not diagnosed carefully, the life time of the patient will not be the greater part a year. Such tumors can quickly prompt passing. Thus, a programmed framework is required for mind tumor identification at a beginning period. In this paper, a computerized strategy is proposed to effectively separate amongst harmful and cancerous free Magnetic Resonance Image (MRI) of the mind. Diverse systems are imposed to isolate tumor. At that point feature set has been considered at each tumor region utilizing Intensity, shape and surface. By then, a popular classification technique called Support Vector Machine (SVM) is imposed by various cross validations on the features set to look at the accuracy of structure introduced in this paper. The new technique approved on a standard dataset, BRATS. The strategy accomplished with average accuracy of 98.2%, area under curve is 0.98, sensitivity of 92.8% and specificity of 98.5%. This method can be utilized to distinguish the brain tumor with much accuracy when contrasted with earlier techniques proposed.
Keywords: Brain Tumor, Pre Processing, Segmentation, K-Means, Morphological Operations, Feature Extraction, Classification, and Support Vector Machine.
Scope of the Article: Classification