Research And Implementaion Of Medical Image 3D Reconstruction System

Medical image 3D reconstruction is a technology which reconstructs the two-dimensional(2D) medical image sequences into stereoscopic effect of three dimensional image by computer visualization technology displayed on the screen, and provides users with a series of interactive functions. At present the latest research results of medical image 3D reconstruction has been applied to medical diagnosis, medical teaching, simulated operation, treatment guiding,etc., especially in the aspect of medical diagnosis, because it can provide real 3D medical image; medical image 3D reconstruction allows doctors to do quantitative and qualitative analysis to their interested lesion area, other tissues and organs from multiple perspectives and multiple levels. It can improve greatly the accuracy of doctors’ diagnosis, thus providing strong support for doctors’ diagnosis and treatment. However, currently some problems still exist in the process of medical image 3D reconstruction, such as unsatisfying reconstruction effect, slow reconstruction speed, stumbling interaction as well as inaccurate image extraction of some lesions or organs, therefore the research of medical image 3D reconstruction still has great significance, and it is also the hotspot in the field of current research of medical image visualization.The main objective of this paper is to achieve a 64 bit medical image 3D reconstruction system software in Windows7 64 bit system, which focuses on how to improve the system throughput, how to optimize the effect of reconstruction, how to speed up the reconstruction and interaction, how to improve the accuracy of image extraction of interested area and so on. In this paper, the medical image 3D reconstruction system has been realized by further studying the medical image DICOM standard, the digital image processing technology and 3D reconstruction technology, as well as combining VTK visualization toolkit and DCMTK software development kit.The main content of this paper are as follows:1、It further studies and introduces the DICOM standard of medical image,DCMTK development kit and VTK visualization toolkit, analyzes the volume rendering technology in detail which plays an important role in 3D reconstruction, and then studies the architecture and the rendering pipeline of graphics processing unit(GPU)systematically.2、It studies the digital image processing technology, and pinpoints the image segmentation technology. In order to extract the cephalic blood vessels accurately, and to improve the reconstruction speed of image data, it proposes an automatic bone removal algorithm and analyzes the effect and efficiency of this algorithm in the experiment.3、It further studies the working principle and features of the graphics processing unit(GPU), and introduces the ray-casting algorithm of volume rendering technology meticulously. Based on this study, it implements the ray-casting algorithm based on GPU acceleration. Furthermore, it compares and analyzes the execution efficiency of this algorithm with the traditional ray-casting algorithm.4、It develops the 64 bit medical image 3D reconstruction system with the combination of VTK visualization toolkit and MFC development platform, and realizes those functions such as compression and decompression images read and display,DICOM file information extraction, one-key bone removal or the head algorithm,volume rendering and interactive module design of transfer function, surface rendering and its attributes module design, some functions of 3D image such as display, rotation,zoom, translation, etc. Last but not least, a large amount of experimental analysis and image display has been done in this paper to test the function effects, execution efficiency and stability of the system.