Magnetic detection electrical impedance tomography with total variation regularization.

摘要: Magnetic detection electrical impedance tomography (MDEIT) is an imaging modality that aims to reconstruct the cross-sectional conductivity distribution of a volume from magnetic flux density surrounding object. The MDEIT in- verse problem inherently ill-posed, necessitating use regularization. most commonly used L 2 norm regulariza- tions generate minimum energy solution, which blurs sharp variations reconstructed image. Consequently, this paper presents total variation (TV) regularization preserve discontinuities and piecewise constancy re- constructed primal dual-interior point method (PD-IPM) employed for minimizing TV penalty in pa- per. proposed validated by simulated data. In comparison with performance tion, results show regularized algorithm produces sharper images has better robustness noise. gularized preserves local smoothness constancy, leading improvements localization MDEIT.