Research On The Algorithm Of Dynamic Induced Current Electrical Impedance Tomography And Imaging Results Based On Physical Phantom

Dynamic Induced Current Electrical Impedance Tomography (ICEIT) is one important branch of Electrical Impedance Tomography (EIT). The main difference between ICEIT and traditional EIT is that ICEIT uses induced current to drive while traditional EIT uses injected current, which can improve the current distribution of the inner part of the image area, and make the measured boundary voltage reveals more impedance information of the inner image area. However, Dynamic ICEIT has a more complicated reconstruction algorithm, needs enormous computation, and requires higher precision of hardware system. Thus, the current image quality of Dynamic ICEIT is not very good.This paper mainly studies the shortcomings of the current ICEIT algorithms and provides some new methods on the basis of old algorithm, which effectively improve quality of the algorithms and obtain better images.My works as followed:1. Study and complete the mathematic model of ICEIT, the formulations, equations, methods of the algorithms of ICEIT, and so on.2. Provide a new method: Iteration Sensitive Matrix Method by studying the current reconstruction algorithms of ICEIT: the Sensitive Matrix Method and the Newton Iteration Method.3. Develop an advanced method: adaptive grouping method in order to resolve the problem that the number of unknown parameters is always great than the number of equations of the algorithm of ICEIT,4. Reconstruct the ICEIT images using the BFGS variable metric method and obtain better images.5. Develop an integrated software experimental tool package on MATLAB.6. Reconstruct images on simulation phantom using the methods of this paper, and compare them with other algorithms, furthermore, reconstruct images with noise signals.7. Reconstruct dynamic ICEIT images on physical phantom. Voltage data come from the dynamic induced current electrical impedance tomography hardware system designed by our group, which has 32 measuring electrodes and the stimulating source is the sinusoidal current. The background of phantom is the NaCl solution with different density. The measured objects are the agar blocks, which have different configuration and are cut into165;cylinders with different shape and area at the transaction, and glass tube or cup, which may be thought as insulator.According the imaging results of computer simulating models, we can find that the reconstructed conductivity distribution is close to the preset objects and background. The reconstruction error is small and the result conductivity value is quite accurate. The imaging objects are clear and have big contrast to the background. The background is even. Further more, our method improves the anti-noise ability.We reconstruct the dynamic images using the data measured from our physical phantom. The BFGS variable metric method is nonsensitive to noise, the result background is even on the whole. The objects in the images are clear. The area of objects is close to that of real ones. The objects are located at the correct position. The iteration sensitive method is more sensitive to noise, the result background is not very even on the whole. The objects can be shown and located at the correct position.In next research, we are to construct a more integrated mathematic model of ICEIT, develop fast reconstruction algorithm, reconstruct images of three dimensions and real / imaginary part, and study on contactless measurement.