Completely Open Mri Magnet Optimization Design

Magnetic Resonance Imaging (MRI) is a kind of imaging technology of biological magnetic nuclear spin, which depended on the development of computer, electronic and superconductivity. According to the composing of MRI system, there are three parts: magnet system, spectrograph system and computer system. Magnet system composed of main magnetic body, gradient coil and radio frequency coil. Main magnet is one of most important and costliness parts. Now, there are three types of MRI magnet: permanent magnet, superconductive magnet and electromagnetic magnet. Now, in medicine applications, most MRI systems are traditional shape whose magnets surround the imaging region. The character of this kind of MRI system is that the imaging region volume is big with a high uniformity of magnetic field, and the imaging speed is rapid. But the main magnetic structure is also big, and the MRI equipment has a high price. This makes MRI system not convenient in popularization in small hospital. A completely open structure of MRI magnet is presented, which is a monohedral magnet structure for MRI. A monohedral magnet has a target region which is outside the magnet structure. A monohedral magnet is situated on one side of the sample, and produces a slice-shaped region with homogeneous magnetic field. The weight and cost of the main magnet is less than the traditional MRI magnet. The function of chose layer can be implemented by changing the displacement of imaging region. With this measure, a series of gradient coil can be leaved out.At the beginning, the thesis introduce the crux of the magnet design of this kind of MRI magnet, and discuss the possible design method. Then indicates the feasibility of this kind of MRI magnet. After that a monohedral magnet for MRI is presented according to the basic theory of magnetic field and the theory of a pair of magnetic dipoles. In order tosatisfy the requirement of the medicine application of MRI, the initial design structure is improved by adding compensation parts. Then a hybrid algorithm of Response Surface Model Methodology(RSM) combined with improved Genetic Algorithm(GA) is presented to optimize the shape of the improved magnet structure. The numerical implementation indicates that the magnitude of the field in target region is greatly enhanced, and the uniformity of the field is also greatly improved. Both of two parameters are close to the requirement of the application of magnetic resonance imaging.