Quantitative analysis of the reconstruction errors of the currently popular algorithm of magnetic resonance electrical property tomography at the interfaces of adjacent tissues. (1st April 2016)

• Record Type: Journal Article
• Authors: Duan, Song; Xu, Chao; Deng, Guanhua; Wang, Jiajia; Liu, Feng; Xin, Sherman Xuegang
• Is Part Of: NMR in biomedicine. Volume 29:Number 6(2016:Jun.)
• Page Start: 744
• Page End: 750
• Publication Date: 2016-04-01
• Subjects: magnetic resonance electrical property tomography -- electrical properties -- tissue interfaces -- errors analysis; Nuclear magnetic resonance -- Periodicals; Magnetic Resonance Spectroscopy -- Periodicals; 574
• DOI: 10.1002/nbm.3522
• Languages: English
• ISSNs: 0952-3480

Abstract

Abstract : This work quantitatively analyzed the reconstruction errors (REs) of electrical property (EP) images using a currently popular algorithm of magnetic resonance electrical property tomography (MREPT), which occurred along the tissue interfaces. Transmitted magnetic fields B 1 + were acquired at 3 T using a birdcage coil loaded with a phantom consisting of various adjacent tissues. Homogeneous Helmholtz was employed to calculate the EP maps by Laplacian computation of central differences. The maps of absolute REs (aREs) and relative REs (rREs) were calculated. The maximum and mean rREs, in addition to rRE distributions at the interfaces, were presented. Reconstructed EP maps showed various REs along different interface boundaries. Among all the investigated tissue interfaces, the kidney–fat interface presented the maximum mean rREs for both conductivity and relative permittivity. The minimum mean rRE of conductivity was observed at the spleen–muscle interface, and the minimum mean rRE of relative permittivity was detected along the lung–heart interface. The mean rREs ranged from 0.3986 to 36.11 for conductivity and 0.2218 to 11.96 for relative permittivity. Overall, this research indicates that different REs occur at various tissue boundaries, as shown by the currently popular algorithm of MREPT. Thus, REs should be considered when applying MREPT to reconstruct the EP distributions inside the human body. Copyright © 2016 John Wiley & Sons, Ltd. Abstract : Magnetic resonance electrical property tomography (MREPT) is a recently developed imaging approach, regarded as a potential tool for the early detection of tumors. Reducing reconstruction errors (REs) is important to improve the MREPT technique. This article analyzed quantitatively REs using a currently popular algorithm. Various REs were observed along different tissue interfaces shown in the figure. The mean relative REs ranged from 0.3986 to 36.11 for conductivity and 0.2218 to 11.96 for relative permittivity with respect to different interfaces.