Permanent magnet for "gradient-free" MRI

点击翻译按钮获取中文摘要

MRI is a well-established non-invasive diagnostic technique used in most modern hospitals that provides high quality images of the entire body. However, there are deficiencies that prevent faster growth and broader application of this valuable technique. Among them are high costs, claustrophobic geometry of the magnet and its substantial weight requiring special site preparation. We propose to address these weaknesses by designing and constructing a compact lightweight magnet specifically for the “gradient free” MRI technique called TRASE. While standard MRI requires time-variable strong gradients of the magnetic field, the TRASE technique, developed with the participation of the applicant, replaces spatial encoding using gradients of the magnetic field with the radiofrequency (RF) field, thus eliminating the magnetic field gradient components, namely the gradient coils and amplifiers used in current MRI systems. The TRASE technique also lessens the high homogeneity requirement for the magnetic field. It is therefore possible to produce inexpensive, light and silent MRI systems that can be made broadly available and used for new imaging applications. We have already collected MR images using magnets that are currently available. We now propose to design and construct a light, “slice selective” magnet comprising 3 coaxial rings that generate a homogenous magnetic field within a narrow cylinder which allows slice selection using non-selective RF pulses. This design will allow slice selection in one direction and 2D TRASE in the perpendicular plane. The design of the magnet is the result of advances in application of genetic algorithms to magnets construction, development of TRASE and our unique magnet design. To design the magnet, we will use Opera 3D software which performs finite element analysis. A Matlab script will provide magnet geometries and block specifications that will be optimized by generic algorithm. This data will be transferred to Opera to calculate the field distribution and provide feedback to Matlab for the subsequent iterations. Whereas current standard MRI magnets weigh thousands of kilograms, the estimated weight of the new magnet for imaging of the human wrist is 10kg. We intend to use an existing 3D printer to fabricate support for the magnetic blocks. The magnetic blocks will be purchased from Newland, China. The homogeneity of the magnet will be measured over a cylinder using a 3D Hall probe attached to an automated robotic arm and analyzed using cylindrical harmonics. Field inhomogeneities will be minimized by correcting the positions of the magnetic blocks and by fine shimming using small compensating magnets to bring homogeneity and distribution of the magnetic field to the target value. The magnet will allow construction of a prototype of the light TRASE MRI system in our laboratory.