Feature Extraction and Performance Evaluation of Classification Algorithms for Liver Tumor Diagnosis of Abdominal Computed Tomography Images

N. Shanila, R.S. Vinod Kumar and N.A. Abin

Liver is the largest internal organ in the human body. Due the heterogeneous nature of liver tissues the hepatic cancer detection is a challenging task. Early detection of primary liver cancer can decrease the mortality rate. Various classification algorithms have been developed for identifying the affected area in liver images. In this paper, Gray Level Co-occurrence Matrix (GLCM) is used to extract classification features in the Computed Tomography (CT) images. This paper is also intended to compare the classification accuracy of various classification algorithms. 10-fold cross validation is used to analyse the performance of the classifiers. Experimental results showed that the k-NN with nearest neighbour 10 achieved the highest accuracy.

Volume 12 | Issue 3

Pages: 82-90

DOI: 10.5373/JARDCS/V12I3/20201169