Study On Three Dimensional Open-style Magnetic Resonance Electrical Impedance Tomography (MREIT)

The electrical characteristic of biological tissues is one of the focuses in themodern biomedical researches. Because the electrical conductivity is different indifferent tissues, and there are quite different values between the normal tissues anddiseased tissues, the images of electrical conductivity distribution is helpful forclinical pathological diagnosis and the exploration on the mystery of human itself.Electrical impedance tomography is an imaging technology by measuring the surfacepotential distribution with the injection of low frequency currents to reconstruct theinternal electrical conductivity distribution of the imaging object. Due to the littleamount of measurement information, the traditional electrical impedance imagingespecially static imaging is facing with the bottleneck of low resolution, which haslimited its clinical application.Magnetic resonance imaging can get the images of multiple slices by computerreconstruction using the magnetic resonance signal produced by the nuclear ofhydrogen atoms in human tissues with the excitation of radio frequency pulse in astatic magnetic field. When an electric current is injected into an imaging object in amagnetic resonance scanner, the magnetic resonance signal will be affected by themagnetic flux density which is induced by the current density. MREIT (Magneticresonance electrical impedance tomography, MREIT) is an imaging modality whichcan measure the magnetic flux density induced by the injected current with a magneticresonance scanner, and using all the internal magnetic flux density data to reconstructthe electrical conductivity distribution. It can provide the exact geometric informationand the internal magnetic flux density of the imaging object, improve the ill-posedcondition in conventional EIT and obtain the electrical conductivity distributionimages with high resolution, which can provide conditions for the clinical applicationof electrical impedance imaging. MREIT is developed into three dimensional spacesin this work, and some image processing methods and current injection modes havebeen put forward especially for open magnetic resonance systems whose mainmagnetic field is vertical to improve the imaging speed and image contrast, whichestablishes the foundation for magnetic resonance electrical impedance imagingtechnology development. 1. The clinical application of three dimensional MREIT on breast cancerdetection is simulated. When developing MREIT from two dimensionalreconstruction to three dimensional reconstruction, it is found that due to the largesize of sensitivity matrix calculation and solution of large equations, it needs a lot ofmemory spaces and takes a long time in conductivity reconstruction. By analyzing thesensitivity matrix, the improved sliced sensitivity algorithm is proposed which canreduce the requirements of computer capability and reconstruction time for3DMREIT, with a high imaging precision and efficiency.2. In order to achieve fast reconstruction for MREIT, based on the forward andinverse problem, a method is proposed for electrical conductivity reconstruction withonly one current injection. Due to single current injection and one component of themagnetic flux density measurement, the scanning time can be reduced and theelectromagnetic interference induced by the electrodes can also be decreased;simulation and phantom experiments have been done to verify the proposedreconstruction algorithm.3. An annular electrode mode is proposed for three-dimensional MREIT in theopen-style MRI systems whose main magnetic field is vertical. Several currentinjection modes in open-style MRI systems were firstly analyzed for MREIT, theannular electrode mode was presented. Simulation experiments were done to test thefeasibility of the proposed annular mode. It is found that with the annular electrodemode, there is similar distribution between the current density and electricalconductivity, which gives magnetic resonance current density imaging more extensiveclinical significances. The simulation results based on a three-dimensional cylindricalmodel and realistic geometry leg model show the feasibility and validity of annularelectrode mode for open MREIT.4. In MREIT, the image quality depends on the SNR (Signal to noise ratio, SNR)of the measured magnetic flux density. In order to improve the image contrast ofMREIT especially under low magnetic field MRI scanners, a post processing methodis proposed. The steps are: gray filling background region, normalizing gray values,equalizing histogram, and stretching gray values. First, by studying the imagingprinciple and operation procedure of MREIT, the necessity of image enhancementwas indicated. Next, an enhancement method for conductivity images was presented.Finally, an experimental validation with a phantom on a0.36T MRI scanner wasperformed. Due to the low SNR of MRI scanners, in the reconstructed conductivity images, most gray values are distributed in a narrow range. While after theenhancement processing, the contrast value is increased from151to356. Thepresented results strongly suggest that after the post processing presented in this work,the grayscale distribution gets more unified in a wide dynamic range, and the imagecontrast is improved significantly.