Research On Preprocessing Algorithms For3D Reconstruction Of Medical Images

Three-dimensional reconstruction convert two-dimensional medical image to volume data and then display three-dimensional image.Three-dimensional reconstruction can enhance our comprehension of organ structure and three-dimensional form of disease. It help to enhance the accuracy of medical diagnose also.In application to medicine, it is hard to establish three-dimensional solid structure and easy to bring distortion of disease position. Three-dimensional reconstruction skill can build three-dimensional image of organ accurately. Provide real-looked three-dimensional medical image, enhance accuracy and science of medical diagnose.Study on three-dimensional reconstruction of medical images has important significance on science and worthiness in practical application.Three-dimensional reconstruction of medical images is a multi-disciplinary subject.It is an important application of computer graphics and image processing in biomedicine engineering.It is related to the subjects of digital image processing, computer graphics and some related knowledge of medical.The main research contents of three-dimensional medical images reconstruction include image acquire, pre-processing,segmentation and extracting tissues of organs, form volume database and constructing three-dimensional models.Preprocessing is the first step and important step of three-dimensional reconstruction system, the result is crucial to segmentation effect and three-dimensional reconstruction system. In this dissertation,key work will be done as follows:1.Research on pre-processing algorithms and segmentation algorithms of three-dimensional reconstruction system, detailing all kinds of the common techniques and a series of experiments were analyzed,pointing out its merits and demerits.In practical applications,appropriate preprocessing algorithm should be selected based on the needs of the system, filter medical images by auto adjusted mid-value filter method to enhance the quality of image.2.Traditional segmenting algorithm need to select subject and ground of every image by user.According to the strong relativity between medical CT image sequence, an algorithm is proposed for the segmentation of medical image.Take the contour of the first image as the edge of the second image and then segmented by region growth after erode the edge. The result shows that this algorithm is high speed and high quality. It is just need to deal with the first image by intervention and to deal with next image automatic.3.Combine the segmentation object and reconstruction algorithm to build an integrated three-dimensional reconstruction system.