The selection of Tikhonov regularization parameter in dynamic electrical impedance imaging / 生物医学工程学杂志

Dynamic electrical impedance imaging is a non-linear inverse problem which is always discretized by finite element method and solved by linear model. Since the coefficient matrix of the linear model is ill-conditioned, the regularization method is used to get stable solution. However Tikhonov regularization is easy to realize, the selection of the regularization parameter is bothering. In this paper, according to the results of a great number of experiments, a new method for determining the regularization parameter is presented which uses the characteristic of the product of the norm of the error and the regularization solution. The reconstructed image using the regularization parameter selected by the proposed method has good spatial resolution.

Image reconstruction in electrical impedance tomography based on genetic algorithm / 生物医学工程学杂志

Image reconstruction in electrical impedance tomography (EIT) is a highly ill-posed, non-linear inverse problem. The modified Newton-Raphson (MNR) iteration algorithm is deduced from the strictest theoretic analysis. It is an optimization algorithm based on minimizing the object function. The MNR algorithm with regularization technique is usually not stable, due to the serious image reconstruction model error and measurement noise. So the reconstruction precision is not high when used in static EIT. A new static image reconstruction method for EIT based on genetic algorithm (GA-EIT) is proposed in this paper. The experimental results indicate that the performance (including stability, the precision and space resolution in reconstructing the static EIT image) of the GA-EIT algorithm is better than that of the MNR algorithm.