[The selection of Tikhonov regularization parameter in dynamic electrical impedance imaging].
Yuan Peng, Yulong Mo
Abstract
Yuan Peng, Yulong Mo
Abstract
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.
OpenAlex reports 3 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
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.
Key concepts: Tikhonov regularization, Regularization (linguistics), Regularization perspectives on support vector machines, Discretization, Inverse problem, Backus–Gilbert method, Applied mathematics, Coefficient matrix