The Best Sensitivity Matrix by Matrix Transformation in Electrical Impendence Tomography

As an advanced visualization technique, the electrical impendence tomography (EIT) can reconstruct the distribution of electrical parameter and thereby visually show the object distribution within a detection field. The EIT reconstruction quality greatly depends on a selected sensitivity matrix, while various matrixes can lead to very different EIT reconstruction results. A large number of efforts to improve the sensitivity matrix have been made to enhance EIT reconstruction quality, but how to construct and select the best matrix in a generally and feasibly way remains unsolved to date. In this paper, we use the matrix transformation method to address the issue, and two types of symmetric and diagonally dominant matrixes are optimally selected to multiply the sensitivity matrix on left and right, respectively. Therefore, the EIT reconstruction quality can generally be improved. The optimality and generalization of the proposed method have been theoretically demonstrated when specially using either Gaussian kernel-based or power function-based matrices, respectively. Experiments validate the proposed method by the EIT reconstruction quality.