No abstract

This paper proposes an automatic system for early detection of liver diseases from Computed tomography (CT) images. The general Computer Aided Diagnosis (CAD) system, including liver diagnosis can be done by segmenting a liver and lesion, extracting features and classify disease whether it is hepatoma or hemangioma. To segment a liver from CT abdominal images histogram analyzer and morphological operation is used. Then to extract a lesion from liver Fuzzy c-mean (FCM) clustering is used. In feature extraction biorthogonal wavelet, Gray-level co-occurrence matrix (GLCM) and fast discrete curvelet transform (FDCT) techniques are used. The textural information obtained was used to train various neural network such as Back propagation Neural Network (BPN), Probabilistic Neural Network (PPN) and Cascade feed forward BPN (CFBPN).The outcome obtained from neural networks are compared with each other to find best combination of features and neural network.

No abstract

Feature selection is an essential one in building high performance classification systems with the maximum classification accuracy. In this paper Particle Swarm Optimization (PSO) hybridized with Sequential Forward Selection (SFS) and Sequential Backward Selection (SBS) algorithm is proposed for improving the performance of the classification system. The feature subsets are extracted from the pattern under classification using First Order Statistics (FOS) combined with the Co-occurrence based features for different distance and degrees. Binary Particle Swarm Optimization (BPSO) is applied to the feature subset. After some iteration the 30 % of the worst particles in PSO is replaced by the best feature subset of SFS and SBS algorithm. The proposed algorithm improves search ability and investigates two types of hybridization (1) PSO-SFS and (2) PSO-SFS-SBS with two options (1) velocity reset of all particles and (2) velocity reset of only worst particles. This hybrid system is applied to liver cancer data to reduce the features and to classify the liver disease as benign or malignant. Liver diseases like Hepatic Cellular Carcinoma (HCC), hemangioma, Focal Nodular Hyperplasia (FNH) and cholangiocarcinoma are classified. The Region of Interest (ROI) is cropped from an abdominal CT. The results obtained from different hybridized feature selection methods are examined. Experimental results show that the proposed methods select the 40 % of features as best features to train the Probabilistic Neural Network (PNN) classifier with insignificant time to categorize the disease to give the accuracy of 96.4 % for data set-1 and 92.6 % for data set-II.

Abstract. Dimensionality reduction of a feature set is a usual pre-processing step used for image classification to improve their accuracy. In this paper an automatic Computer Aided Diagnostic system (CAD) is proposed for detection of liver diseases like hepatoma and hemangioma from abdominal Computed Tomography (CT) images using an evolutionary approach for feature selection. The liver is segmented using adaptive thresholding. Histogram analyzer is used to fix the threshold and morphological operation is used for post processing. Rules are applied to remove the obstacles. Fuzzy c-Mean (FCM) clustering is used to extract the lesion from the segmented liver. Auto covariance features are extracted from the segmented lesion. The Binary Particle Swarm Optimization (BPSO) is applied to get the best reduced feature set. The textual information obtained after feature reduction was used to train Probabilistic Neural Network (PNN). The results obtained from different transfer functions are analyzed and compared.