Segmentation and classification of liver ct images for tumor detection

Abstract

The liver is one of the essential organs positioned inside the abdominal newlinepart of the human body. The purpose of the liver includes the production of biles, newlinepurification of blood, storing of vitamins and minerals and metabolic function. newlineIn addition, to create nutrients and to metabolize the drugs in the form of glycogen newlineliver is considered to be an important organ. Due to these reasons, it is highly newlinesignificant to defend the liver from the tumor. The tumor is an abnormal cell newlinegrowth. Depending on the mass of the tissue, it can be classified as primary liver newlinetumor and secondary liver tumor. Several imaging modalities such as Nuclear newlineMedicine Imaging (NMI), Ultrasonography (US), X-Ray, Computed Tomography newline(CT), Magnetic Resonance Imaging (MRI), and Positron Emission Tomography newline(PET) are used by the experts for detecting both liver and tumor portions. newlineHowever, CT modality can show detailed images of the human body such as bone, newlinemuscles, abdomen, heart, lung and kidney. The other modalities have the newlinelimitations such as high cost, long scanning time, deprived tumor localization, and newlinerequires skilled trainers. Considering these reasons, CT is extensively used in liver newlinetumor diagnosis. newlineThe segmentation and classification of liver tumor from abdomen newlineimages has been a demanding task in recent days. Various segmentation newlinemethodologies such as manual, semi-automatic, and fully automatic segmentation newlineare developed for segmenting the image. The segmentation results afford newlineby the experts may vary from patient to patient and also for the same patient newlineat different times. Also, it takes more time to diagnose the tumor as more slices newlineare needed for a better diagnosis. Thus the problems raised by manual newlinesegmentation are overcome by developing a semi-automatic segmentation newlinemethod. In semi-automatic method, human intervention is needed to locate the newlineseed point within the roughly selected liver region newline