The Analysis And Implementation Of Hip Joint Feature Extraction System

The medical image processing can effectively guide the clinical diagnosis and promote the development of the medical research for its ability to digital, modelized and specific the medical information. Extracting the feature of the medical images, and then measure and calculate the specified indicators according to the extraction data is the basic step.Firstly, the paper explains the requirement analysis and module division of the system which is based on the extraction of the hip feature. The critical modules are the contour extraction, the feature extraction and the three-dimensional reconstruction of the target bone. Then, analyzing the target images and the evolution trend of bones in the image sequences based on the bone module to get the specific cases, which have similarities in the sequence frames and outline differences in some areas. Because it is not convenient to extract all features of many targets, so the paper adopts the divide and conquer method to extract the outline of femur at the beginning, then the acetabular bone, which can obtain more effective extraction data.Secondly, calculating the feature indicators and reconstructing the model of bones according to the outline data. The first step is to analyze the indicators of the femur and acetabulum, including the calculation of center points of the femur bottom, the greater trochanter neighborhood and the femoral head, and the outline dissection of the plane defined by these centers. The second step is to acquire the indicators of femur features depend on the data of centers and dissection, while the acetabulum indicators could be calculated after the three-dimensional reconstruction, the reconstruction can not only three-dimensional the two-dimensional outline to cause the intuitive feeling, but also can get the acetabulum indicators by adding the function of dissection of any planes in the model. Moreover, the reconstruction also supports operations of rotating, zoom and so on, which provides the convenience of analysis and observation of images from any aspects.Finally, the paper proposes the direction and thoughts about the future work, which will do guiding meaning to the deep research and exploring. Seventy-three figures, one table, seventy-one references.