An approach on liver medical image segmentation based on quad tree

In this paper we present a medical graph cuts based quad trees segmentation method. Our method is a combination of quad trees decomposition and curve fitting method that improved the shortcomings of the original method. Consequently, it has the following advantages: (1) It has the ability to fill holes that modify the original method defect. (2) Our method guarantees continuity and lead to smooth contours. From experiment results we can know the feasibility and advantages of the method; it is suitable for the actual liver segmentation. Our method also can be the foundation of the subsequent diagnosis.