Gallbladder Segmentation in 2-D Ultrasound Images Using Deformable Contour Methods

摘要: Segmenting the gallbladder from an ultrasonography (US) image allows background elements which are immaterial in diagnostic process to be eliminated. In this project, several active contour models were used extract shape of gallbladder, both for cases free lesions, and those showing specific disease units, namely: lithiasis, polyps, anatomical changes, such as folds or turns gallbladder. First, histogram normalization transformation was executed allowing contrast US images improved. The approximate edge found by applying one like motion equation, a center-point model balloon model. An operation adding up areas delimited determined contours also more exactly images. Then, fragment located outside eliminated image. tests conducted have shown that 220 area error rate (AER) amounted 16.4%.