Key techniques research in computer-aided hepatic lesion diagnosis system based on multi-phase CT images

Computer-aided diagnosis (CAD) of liver diseases as an early non-invasive diagnosis is of great significance. This paper presents an automated diagnostic system for liver disease based on multi-phase CT images. The region of the liver is first extracted from a CT image using improved watershed algorithm. After the registration of liver regions, which uses the SIFT algorithm, the operation of extracting the ROI based on Gabor wavelet transformation would be followed. Besides using image texture metric as the feature vector, we also designed a temporal and sacttergram-based lesion enhancement pattern descriptor to quantify the different lesions. Then, in the designing of classifier module, we convert a 4 classes classifying problem into 3 binary classify problems by using artificial neural network. Finally, we obtained the best classification accuracy of 0.9797, 0.9851 and 0.9753 for normal-abnormal, cyst-otherdisease and carcinoma-haemangioma sub problems respectively.