The Study Of Tumour Functional Image Based On Bioimpedance Technology

Electrical Impedance Tomography(EIT) is a new type of medical imaging technique developed in recent years. In the biology, different tissues have different electric parameters, such as conductance and capacitivity. When a small altering current is injected to an object, corresponding voltages can be measured via some electrodes attached to the object. Using above data, an approximation distribution for the internal impedance can be obtained by some EIT imaging reconstruction algorithms. Bioimpedance technology has some advantages over existing imaging methods, it is of non-invade, low-cost and continuous-inspecting, it can offer a functional image, which has a physiological and pathological information. In the delitescence of some disease, the impedance of the pathological tissues may be changed and these changes can be reflected by the electrical characteristics in the biological tissues. So EIT can be used to study some diseases. In clinical medicine, this technique can be used to preliminary discovery and diagnoses of tumour, falling sickness and other diseases. Two main parts are included in this paper. One is the electrical impedance tomography algorithm based on RBF artificial neural network. The other is the tumor detection system based on impedance measurement.To study the EIT imaging reconstruction algorithm, First, a detailed process of EIT forward problem solving based on Finite Element Method is given. The forward problem's solving process involves the finite element theory, the variation of Laplace equation, and the finite element mesh partition way of regional models, etc; Second, does a specific analysis and detailed studies in researching the inverse problems of EIT and the ill-conditioned character. Third, give specific expositions to the algorithm principle of EIT based on RBF artificial neural network, RBF network architecture design, object model and sampling design, network training methods, image reconstruction process. Last, simulate the imaging algorithm process and analyze the factors of impacting image reconstruction effect.To solve the simple genetic algorithm's premature convergence and slow convergence phenomena, we learn from the natural world's common sexual reproduction and"niche"phenomenon, integrate gender, age and diploid coding method, import the niche selection technology, present the niche genetic algorithm based on sexual reproduction. To simulate the tumor(eg: breast cancer, lymphoma) detecting system through establishing simplified model, using two genetic algorithms to locate the tumor, by contrast, the effect of tumor location of niche genetic algorithm based on sexual reproduction was proved.