International Journal of Advanced Computing and Communication Systems

Classification of Tumors using PCA and SVM from MRI

Palraj K, Srividya college of Engg & Tech, Engineering college, Virudhunagar-626005, Tamil Nadu, India.

Kalaivani V, National Engineering college, Kovilpatti-629005, Tamil Nadu, India.

International Journal of Advanced Computing and Communication Systems

Received On :

Revised On :

Accepted On :

Published On :

Volume 06, Issue 01

Page No :009-014

Abstract

Classification of tumor tissues in magnetic resonance images (MRI) is a significant task but is relatively time overwhelming while achieved manually by proficient. Automating this progression is difficult due to the high miscellany in exterior of tumor tissue in different patients, and in many cases, resemblance between tumor tissues and normal tissues. This paper presents an automatic technique for classification of tumor tissues in MRI. After pre-processing and elimination of the regions that do not have useful information, it creates a projection on PF affected tumor region, when pre-processing and elimination of the regions that do not have useful information was done. In this work feature extraction from MR Images is agreed by Daubachies(DAUB-4) Wavelet technique. Using Principal Component Analysis (PCA), to choice the unsurpassed features for classification. Support Vector Machine (SVM) takes the selected features as an input from PCA. This enhances the tumor boundaries more and is very fast when compared to many other classification algorithms.

Keywords

homomorphic filtering; canny edge detector; high boost filtering; pratt’s figure.

Cite this Article

Palraj K and Kalaivani V, “Classification of Tumors using PCA and SVM from MRI , ”International Journal of Advanced Computing and Communication Systems, pp. 001-008, March 2019.

Copyright

© 2019 Palraj K and Kalaivani V. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.