Premium
Multi‐echo GRE ‐based conductivity imaging using K alman phase estimation method
Author(s) -
Ryu Kanghyun,
Shin Jaewook,
Lee Hongpyo,
Kim JunHyeong,
Kim DongHyun
Publication year - 2019
Publication title -
magnetic resonance in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.696
H-Index - 225
eISSN - 1522-2594
pISSN - 0740-3194
DOI - 10.1002/mrm.27376
Subject(s) - extrapolation , imaging phantom , phase (matter) , nonlinear system , conductivity , kalman filter , echo (communications protocol) , algorithm , nuclear magnetic resonance , computer science , physics , mathematics , artificial intelligence , mathematical analysis , optics , quantum mechanics , computer network
Purpose To obtain in vivo electrical conductivity images from multi‐echo gradient‐echo (mGRE) sequence using a zero‐TE phase extrapolation algorithm based on the Kalman method. Methods For estimation of the zero‐TE phase from the mGRE data, an iterative algorithm consisting of a combination of the Kalman filter, Kalman smoother, and expectation maximization was implemented and compared with linear extrapolation methods. Simulations were performed for verification, and phantom and in vivo studies were conducted for validation. Results Compared with the conventional method that linearly extrapolates the zero‐TE phase from the mGRE data, the phase estimation of the proposed method was more stable in situations in which nonlinear phase evolution exists. Numerical simulation results showed that the stability is guaranteed under various nonlinearity levels. Phantom study results show that this method provides improved conductivity imaging compared with the conventional methods. In vivo results demonstrate conductivity images similar to spin echo–based conductivity images with the added benefit of the acquisition of susceptibility images when using mGRE. Conclusion The proposed method improves zero‐TE phase extrapolation, especially in regions of nonlinear phase evolution. Improved conductivity imaging using mGRE can be performed.