Application of constrained coefficient fuzzy linear programming in medical electrical impedance tomography

Abstract

Electrical impedance tomography (EIT) is an imaging technique that realizes the image reconstruction of conductivity distribution in the field. The existing EIT algorithms ignore the hidden fuzzy features during imaging, making the EIT technique exhibit a high degree of uncertainty, imprecision, incompleteness, and inconsistency in the actual use process, resulting in a low spatial resolution of the reconstructed images. In order to solve this problem, we introduce fuzzy linear programming into EIT imaging. On the basis of analyzing the fuzzy features of EIT in detail, a new model of fuzzy optimization is built up, whose optimal solution is obtained by using constrained coefficient fuzzy linear programming. To this end, we devise two types of simulation experiments to verify the performance of the optimization algorithm. Experimental results prove that compared with the traditional Tikhonov regularization algorithm, the correlation coefficient of the reconstructed image of the proposed algorithm is higher, the relative error value is smaller.

Keywords

• electrical impedance tomography
• fuzzy linear programming
• fuzzy features
• constraint coefficients
• image reconstruction
• uncertainty
• inverse problem
• tikhonov regularization
• lung cancer detection
• model