| There are two main kind of clinical medicinal imaging technology. One is the structural imaging, and the other is the functional imaging. The former includes X-ray, CT, MRI and ultra-sound imaging. And the latter includes the radionuclidic imaging(PET), fMRI, Magnetoencephalography(MEG) and EIT. And now there is a urgent demand in the clinical application for a portable real-time monitoring functional imaging system on the bedside. All the mentioned imaging method can not satisfy this demand but EIT, but EIT employs a pair of electrodes to inject current into the measured objects, the low-conductivity characteristic of the skull layer limits its application in the brain disease functional imaging. Because the magnetic field can pass through the low-conductivity skull layer almost without attenuation, in this paper a contactless functional imaging method: Open Magnetic Induction Tomography is introduced. The research content includes following four parts:①Forward problem of the open magnetic induction tomography. There are two parts of the forward problem research. A -φmethod based on FEM is employed to study the 2-D eddy current distribution. Referring the separated-layer idea in the geologic research, an assumption of multi-layer biological tissue model is presented. And a Gauss-Legendre numerical integral based on the region-separated idea is employed to calculate the forward problem of the open magnetic induction tomography.②Inverse problem of the open magnetic induction tomography. A modified GA is employed to calculate the inverse problem. The modification of GA includes the following aspects. The initial population is generated according to the generalized Humming distance limitation. The elitist remaining method and the ranking algorithm based on pressure difference are used in selection. The two crossover individuals are selected according to the relative partner method. And gene position of the single-point crossover is located referring the limitation of the active crossover region. The crossover operation probability is adjusted according to the fitness value, and the variation operation probability is adjusted according to the iterative number. The simulation result indicates: a system equips a sensor with 10mm radius can reconstruct a conductivity abnormal region of two times larger as the sensor's size within 20mm under the top surface.③Measurement system research. All circuits and their designing principles are introduced. To simulate the human head, a multi-channel measurement system is constructed.④Experiment research. There are two steps of the experiment research. One is the performance measurement of the MIT system including sensitivity, stability and the noise level. The result indicates that the sensitivity of phase is higher than that of the magnitude, so the phase data of the signal is recorded as the source of the imaging. The phase sensitivity of the system is 0.185~o / S /m . In order to improve the stability of the system and minimize the phase noise, a reference channel is employed. The phase noise range is±0.05°. The other is the water tank experiment. The results of the frequency-changing experiment indicate that the imaging quality can be improved by increasing the exciting frequency. The depth experiment indicates that our system can detect conductivity abnormal region within 20mm under the surface. And the imaging result of the agar model covered by a pig bladebone indicates that the magnetic induction tomography can detect conductivity distribution under the skull. And these experiments construct a base for the future clinical brain disease monitoring.
|
PDF Full Text Request