The Research For Computer-aided Diagnosis Of Fatty Liver Based On Liver Ultrasound Images

With the advent of the era of large data, the amount of medical image data and the dimension of data become more and more abundant, thus forming a huge medical digital image library, such as liver B ultrasound images. How to make the image observer quickly identify the image of a disease (such as the image of fatty liver) from a large number of liver B ultrasound images, thereby reducing the search range, while ensuring a certain accuracy, greatly reducing the observer complicated work Quantity, is a very meaningful topic.First of all, this study collected a large number of liver ultrasound images, collected from the ultrasound image is the first People’s Hospital of Xichuan County, the clinical image, and by the image of the medical imaging doctors were classified to ensure that the number of data sources and quality. Based on the collection of images, the basic principles of liver ultrasound examination were also studied, and a typical ultrasound image section was identified. At the same time, it also describes the medical standards of fatty liver discrimination, as a prerequisite, proposed the use of computer-assisted fatty liver diagnosis criteria. Secondly, this paper introduces the preprocessing method of image, and selects the typical local region of the image by manual selection. In this study, the idea of using texture features to identify fatty liver was determined. Gray scale cooccurrence matrix (GLCM) method was used to express the texture features in the image. Meanwhile, this study used the gray level co-occurrence matrix The covariance matrices are analyzed, and the covariance matrices are analyzed. The matrices are reduced by using the null space method. Thirdly, this study extracts the eigenvalues from the gray-level co-occurrence matrix to reflect the features of the texture and form the vector library. Each vector in the vector library represents the texture information of an image.Because of the different parameters of the gray level co-occurrence matrix, a variety of eigenvector templates are formed, which are divided into normal liver vector and fatty liver vector. Finally, compared with the Euclidean distance between the image feature vector and the feature library vector in the test set, the class of the nearest Euclidean distance image is the liver image category. In view of the limitation of the Euclidean distance method, this paper applies the KNN method in data mining classification technology to this paper. Through a large number of experiments, the effective filtering method can solve the problem and achieve the accuracy of 86.78%.The research results have provided great theoretical value and practical significance, improving the doctor’s work efficiency and accurateness of diagnosis for the rapid screening of fatty liver image in a large number of liver ultrasound images.