Research And Development Of The Software For Three-dimensional Visualization Of Tumor Images, Radiotherapy Target Volumes And Organs At Risk

Medical image three-dimensional visualization is a computer graphics technology which uses a series of two-dimensional medical images of cross-sections of the human body tissues and organs to render their three-dimensional surfaces(surface rendering) or three-dimensional structures (volume rendering). The rendered three-dimensional medical images can provide doctors with intuitive and comprehensive information about the lesion and normal tissue, also can correctly grasp the shape of the intuitive and surrounding tissues, organs and spatial location and distribution information, can visually evaluate the shape of the lesion, surrounding normal tissues and organs and the relative position.The article refers to three-dimensional visualization of tumor medical images such as CT, MRI and PET as well as tumor radiotherapy target volumes and organs at risk(OARs). This paper chooses Marching Cubes algorithm and Ray Casting Composite algorithm to realize the tumor images three-dimensional visualization. Meanwhile, chooses Delaunay triangulation algorithm to realize the tumor radiotherapy target volumes and OARs three-dimensional visualization.Surface rendering method refers to the reconstruction of the surface. It is extracted from the three-dimensional data level for surface fitting and can effectively draw three-dimensional body surface, also can use a variety of technologies to accelerate the speed. But the render result lacks the internal information expression, and takes a lot of time to reconstruct surface accurately. This shortage really limits the medical image visualization development. Volume rendering method is to process every voxel in three-dimensional data field, including calculate the opacity, color and other properties of each voxel. The method is studied direct light through the body data field and the relationship of the voxels and did not to construct the middle surface. Thus voxels can retain many of the details information, but the rendering process takes a lot of time. Delaunay triangulation method is a three-dimensional reconstruction of two-dimensional contour sequence, and the program running time is short and the operation is stable.The system is developed by object-oriented software system design methods, ITK, VTK and FLTK. We designed and developed a subsystem for three-dimensional visualization of tumor images, radiotherapy target volumes and organs at risk, based on our framework of the molecular imaging guided intelligent radiation system. The main contents are as follows:(1) Visualize the tumor images to three-dimensional images, which are intuitively and stereoscopically displayed on the computer screen.(2) Visualize the sketched tumor radiotherapy target volumes and OARs. Thus doctors can directly observe organs of the structure, shape and size, as well as the relative position of various organs.(3) A three-dimensional interactive module is designed to realize the following functions:the organ selection, surface rendering and volume rendering switchover, transparency adjustment, surface rendering threshold adjustment and color editing.