Liver CT enhancement using fractional differentiation and integration

In this paper a digital image filter is proposed to enhance the Liver CT image for improving the classification of tumors area in an infected Liver. The enhancement process is based on improving the main features within the image by utilizing the Fractional Differential and Integral in the wavelet sub-bands of an image. After enhancement, different features were extracted such as GLCM, GRLM, and LBP, among others. Then, the areas/cells are classified into tumor or non-tumor, using different models of classifiers to compare our proposed model with the original image and various established filters. Each image is divided into 15×15 non-overlapping blocks, to extract the desired features. The SVM, Random Forest, J48 and Simple Cart were trained on a supplied dataset, different from the test dataset. Finally, the block cells are identified whether they are classified as tumor or not. Our approach is validated on a group of patients' CT liver tumor datasets. The experiment results demonstrated the efficiency of enhancement in the proposed technique.