Detection of Liver Cancer using Optimized Techniques in CT Scan Images

Detection of liver cancer using computed tomography (CT) scan images is a crucial task in clinical practices. In this paper, we have proposed a novel method for segmenting the liver cancer based on the Optimized techniques such as particle swarm optimization (PSO), Darwinian particle swarm optimization (DPSO) and fractional-order Darwinian particle swarm (FODPSO) algorithms using CT scans. The algorithm was tested in a particular slice from a series of 40 real-time images having cancerous affected regions collected from the different subjects at IMS and SUM Hospital, India. The proposed method includes pre-processing, segmentation, feature extraction and classification stages. Initially images were segmented with optimized techniques, then various statistical and morphological features were extracted from the segmented images. The feature set was then classified into two types of liver cancer i.e. hepatocellular carcinoma (HCC) and metastatic carcinoma (MET) using decision tree (C4.5) classifier. The method effectively segmented the lesion structure in FODSPO process with accuracy of 97.5% which is better than PSO and DPSO methods. The obtain results confirmed the superiority of FODSPO technique with C4.5 classifier for identifying the liver cancer.