Fatty Liver Diagnosis Of B-mode Ultrasonography Based On The Complexity Analysis

Fatty liver contains an excessive amount of fat and the normal healthy liver tissue is partly replaced with areas of unhealthy fats. If it is left untreated, fatty liver may be associated with or may lead to inflammation, fibrosis, cirrhosis and possibly cancer. The B-mode ultrasound is widely used to diagnose fatty liver. But it relies on the doctor's observation and experience. So it is subjective and may be unreliable. With the development of texture analysis technology, it is reasonable to diagnose fatty liver by texture recognition of B-mode ultrasonic image. As a nonlinear method, complexity analysis is widely used to study the EEG, ECG and CT images. So it can be used in research of the confusional information of the image.This study focuses on the recognition techniques of the fatty liver based on complexity analysis. Fatty liver may be associated with or may lead to inflammation of the liver, and the texture of the B-mode ultrasonic image changes. Firstly, B-scan ultrasonic images are captured from Huaxi Hospital of Sichuan University. Secondly, the regions of interest are localized. During the feature extraction procedure, mean intensity ratio of the near field and the far field(MIR), the approximate entropy(ApEn), the Kc complexity and the gray level co-occurrence matrices features are extracted. Then, C-means clustering algorithm and back-propagation (BP) artificial neural network are employed respectively to classify these feature vectors. There are totally 130 patients studied in the experiments. 55 of them are normal liver, and the other 75 are fatty liver in middle or serious degree. When use the MIR, the ApEn, the Kc complexity as the feature vector, the images can be clissfied in two...