الفهرس 
AbstractOnly 14 pages are availabe for public view
 |  | from | 77 |  |  |
| | | from | 77 | | |
Abstract
Chronic liver disease (CLD) is an oncogenic disease and if not treated the most probable outcome is hepatocellular carcinoma (HCC). Detecting cirrhosis in an early stage is the key for the successful treatment. The detection of abnormalities in ultrasound (US) liver images is a tedious task. Consequently, there are solid requirements for the development of computer aided diagnosis (CAD) systems, which have the ability to help radiologists to take the correct decision. The overall objective of this thesis is to develop a novel CAD system for diagnosis of liver cirrhosis using ultrasound images. The proposed system consists of segmentation, feature extraction, feature selection, and classification steps.
In this work, we cropped region of interest (ROI) with 128128 pixels by manually segmented, then features are extracted from the segmented ROI. A combination of different features; morphological, signal processing and statistical features are used. It is then followed by a classification step to determine whether the ROI is normal or abnormal. The classification accuracy rates are calculated using a 10-fold cross-validation study. A correlation-based feature selection (CFS) is used resulting in better accuracy predictions. The results showed that Support vector machine (SVM )and K-nearest neighbor(K-NN )classifiers achieved higher performance with the combination of the wavelet and curvelet feature vectors than other feature extraction methods, with an accuracy rate of 99.31%.
The result of this Thesis has been accepted for publication as referred to in reference [73].