Method of edge energy template for detection of ultrasonic based fatty liver

In the current medical world, B-ultrasound is usually utilized as the detection technique of fatty liver. However, because of the image resolution at B-Ultrasound is limited, the technique is inadequate to after quantitative data to help doctors make decisions and judgments. Therefore, doctors instead make decisions based on their precedent experiences, which are arguably not precise in many cases. For the sake of accurate judgment of fatty liver, it is crucial to combine doctor´s personal judgment and the method of image processing which could extract more accurate features from the B-Ultrasound. Methods of existing image processing mostly based on texture analysis. For example, gray level co-occurrence matrix(GLCM), which calculate quantities and a large number of parameters, is largely complex, and not stable in terms of its parameter. This paper put forward a method based on the use of edge energy template, which was used for processing and calculating B-Ultrasound image of liver with only very few parameters. The method was very expressive, and the result is simple to calculate. One of the creationary ideas for this method is to partition the image into different templates, it helps to reduce redundant calculation, and will not lose any information regarding the pathology. It also designs algorithm of filteration, and improved method of extraction of Region of Interest (ROI) and partition of near-far regions.