Study Of DICOM Medical Image Processing And Three-Dimensional Visualization Technology

With the rapid development of increasingly sophisticated medical imaging equipment and medical technology,the amount of medical image data has increased dramatically.Under the current backdrop of massive image data,many traditional medical image processing algorithms are difficult to meet the new requirements,real-time and efficiency.Therefore,the study of medical image processing algorithms is a progressive research focus.In this paper,most commonly used medical image processing filtering and segmentation operations have been studied and improved.And for the limitations of the traditional two-dimensional reading of clinical piece,three-dimensional medical image visualization techniques were analyzed,and one of the key technologies to accelerate reconstruction has been improved.Since the vast majority of the medical image data are stored in DICOM file format and transmission of DICOM standard,this paper first studied and implemented a fast DICOM file parsing and conversion methods to facilitate follow-up study.Medical image extremely susceptible to noise interference during imaging,so directly use in the case of clinical diagnosis may due to misdiagnosis caused by image blurring.In order to improve the quality of medical images,this paper conducted deep research on medical images filtering.For medical images,the edge feature is often more important information.As traditional filtering method easily lead to the problem of blurring edges of the image information while suppressing noise,this paper conducted intensive studies on anisotropic diffusion model which based on partial differential equations.On the basis of complex diffusion model based on median filtering,termination criteria for the iterative algorithm has been improved anisotropy,iteration termination criterion taking into account the improved image de-noising and edge losses to becapable of flexible choice,and get the best filtering results by the computer automatically.Compared to the original manual identification method,the improved algorithm can obtain more reliable results at the same time,saving labor cost and time overhead.Clinical diagnosis often requires a specific organization of medical images targeted observation and research.In this paper,several mainstream level set segmentation algorithm were compared with the realization.We analyzed the key performance of these methods,and achieved a framework based on medical image processing filter,the level set method of segmentation and 3D visualization.In the medical image visualization process,the speed of three-dimensional reconstruction has been hampered bottleneck of the technological development.This paper implements the Marching Cubes algorithm and Ray Casting algorithm,both the pros and cons and application scenarios are compared and analyzed and summarized.For handling the problem of a large number of redundant triangular facets after the Marching Cubes algorithm processing,an improved quadric error edge collapse algorithm was proposed to increase the speed of three-dimensional reconstruction.The improved algorithm makes full use of the geometric information around the mesh vertices and differences between equilateral triangles with simplified Triangles been regarded as weights to achieve a complete simplified mesh quality in a short period of time.Finally,after the reconstruction volume rendering,the internal structure of the organization is easily obscured by the surface information.This paper based on VTK toolkit implements an interactive arbitrary planar cutting algorithm to extract and display any desired setting tangent plane based on artificial interactive tangent plane to achieve a true all-round three-dimensional visualization.