Renal Cell Carcinoma Nuclear Grading Using 2d Textural Features for Kidney Images
Cancer identification system is proposed based on the features present in the kidney images. Different algorithms such as CLACHE (Contrast limited adaptive histogram equalization), GLCM (gray level Co-occurrence matrices) and SVM (support vector machine) algorithm are used for the identification of cancer. CLACHE algorithm is used for the enhancement of the image. GLCM algorithm is used to improve the overall accuracy of the system and to extract the textural features. SVM algorithm is used to classify the different grading levels to identify the cancer present in the image. Images that are acquired for the identification of cancer are noisy. Noise is removed by the ROI extraction. Then the images are enhanced using CLACHE algorithm. Once the images are enhanced, features are extracted using GLCM. 21 textural features are extracted. Out of the 21 features extracted two best features are selected. The two best features are compared with the trained features for the increase in the accuracy of the system. After that based on the features different grading levels are obtained for the identification of the cancer. Grade 1 indicates the presence of cancer in starting stage, grade 2 indicates the presence of cancer in the moderate stage, grade 3 indicates the presence of cancer in the mild stage, grade 4 indicates the presence of cancer in the severe stage. In this study, 2D textural features are extracted and using these extracted features cancer identification is done which improves the overall accuracy of the system.
Published by: Jayashree G. R, Dr. K. M Ravi Kumar, Ravi Kiran .R
Author: Jayashree G. R
Paper ID: V3I3-1377
Paper Status: published
Published: May 25, 2017
Full Details