| Magnetic Resonance Imaging (MRI) is a noninvasive technique for imaging the internal structure of the body. Comparing to other medical imaging techniques, MRI has no radiation damage, can do image multi-directionally and multi-parametrically and can get very high contrast soft tissue image.The MRI is different from the other medical imaging modalities, MRI data acquisition procedure must first fill the Fourier space (k space), and then the image can be obtained by Fourier transform. Actually after the Fourier transform the obtained data is complex data, and the magnitude image for diagnosis is got from the intensity information from those complex data. In fact the complex data also include the phase information; there are many different way to use the phase information, including phase contrast magnetic resonance imaging (PC-MRI) and susceptibility weighted imaging (SWI) methods. As the PC-MRI can be used to measure the flow, so it has enormous potential in study of human hemodynamics, cardiac function, blood vessel elasticity, which is also the focus of my study. Because the PC-MRI methods is not fast enough and the large data of the PC-MRI image is not easy for doctors to diagnosis, currently the clinical application of the PC-MRI is not yet widespread. The aim of this paper is trying to solve the problem of the PC-MRI methods to promote the clinical application of this method.My works in this thesis include the following:(1) The PC-MRI imaging procedure simulation. Comparing to 1 dimensional MR signal simulation, a 2 dimensional flow phantom MR signal simulation has been applied in this paper. This simulation includes the RF pulse, gradient system and nuclear spin relaxation. Under the assumption that the spin flow forward, the simulation can get the phase contrast image. This simulation is not just simulating the way of MR scanner got the image, but also simulating the way to fill the K space. So this simulation can apply a way to optimaze the parameters of the PC-MRI and the data for MRI reconstruction research. By PC-MRI stationary and flow experiment validated the simulation is very effective.(2) Optimize the PC-MRI sequence to make the measurement of blood flow more efficient. Our research is analysis the RF spoiling and gradient spoiling to optimaze the PC-MRI sequence. First theoritically analysis the RF spoiling and gradient spoiling, then using simulation, phantom experiment and volunteer expriment to validate the optimazed RF spoiling and gradient spoiling.(3) Establish a CPU-based medical image three-dimensional visualization system. According to the demands of clinical application a CPU-based medical image three-dimensional visualization system has been established. This system can run on personal computing system to achieve high quality three dimensional image by optimize the volume rendering algorithm for the CPU structure,(4) Three-dimensional medical image visualization algorithms optimization. In order to enhance the visualization quality of the three-dimensional image, this study applied a novel two-dimensional transfer function and a new reflectance model, and a new way of surface rendering by using the ray-casting algorithm.(5) Accomplish the four-dimensional PC-MRI data visualization methods. Visualize the four-dimensional PC-MRI data can easily observe the blood flow in human as a function of time. By mapping different flow direction and flow rate to different colors and brightness, the visualization result of the blood flow information can be obtained. Then fuse the flow and magnitude visualization result can give the operator a convenient way to observe the four-dimensional PC-MRI data.My research of this dissertation has both the theoretical and practical significance, which not only provide a PC-MRI sequence optimization method also give a solution to visualize the four-dimensional PC-MRI image. This research can be helpful to diagnosis and preclude cardiovascular disease, if the research result been adopted clinically. |
PDF Full Text Request