Medical College of Wisconsin
CTSICores SearchResearch InformaticsREDCap

MREIT experiments with 200 µA injected currents: a feasibility study using two reconstruction algorithms, SMM and harmonic B(Z). Phys Med Biol 2012 Jul 07;57(13):4245-61 PMID: 22684125 PMCID: PMC3381422

Abstract

Magnetic resonance electrical impedance tomography (MREIT) is a technique that produces images of conductivity in tissues and phantoms. In this technique, electrical currents are applied to an object and the resulting magnetic flux density is measured using magnetic resonance imaging (MRI) and the conductivity distribution is reconstructed using these MRI data. Currently, the technique is used in research environments, primarily studying phantoms and animals. In order to translate MREIT to clinical applications, strict safety standards need to be established, especially for safe current limits. However, there are currently no standards for safe current limits specific to MREIT. Until such standards are established, human MREIT applications need to conform to existing electrical safety standards in medical instrumentation, such as IEC601. This protocol limits patient auxiliary currents to 100 µA for low frequencies. However, published MREIT studies have utilized currents 10-400 times larger than this limit, bringing into question whether the clinical applications of MREIT are attainable under current standards. In this study, we investigated the feasibility of MREIT to accurately reconstruct the relative conductivity of a simple agarose phantom using 200 µA total injected current and tested the performance of two MREIT reconstruction algorithms. These reconstruction algorithms used are the iterative sensitivity matrix method (SMM) by Ider and Birgul (1998 Elektrik 6 215-25) with Tikhonov regularization and the harmonic B(Z) proposed by Oh et al (2003 Magn. Reason. Med. 50 875-8). The reconstruction techniques were tested at both 200 µA and 5 mA injected currents to investigate their noise sensitivity at low and high current conditions. It should be noted that 200 µA total injected current into a cylindrical phantom generates only 14.7 µA current in imaging slice. Similarly, 5 mA total injected current results in 367 µA in imaging slice. Total acquisition time for 200 µA and 5 mA experiments was about 1 h and 8.5 min, respectively. The results demonstrate that conductivity imaging is possible at low currents using the suggested imaging parameters and reconstructing the images using iterative SMM with Tikhonov regularization, which appears to be more tolerant to noisy data than harmonic B(Z).

Author List

Arpinar VE, Hamamura MJ, Degirmenci E, Muftuler LT

Author

Lutfi Tugan Muftuler PhD Associate Professor in the Neurosurgery department at Medical College of Wisconsin

MESH terms used to index this publication - Major topics in bold

Algorithms
Animals
Dogs
Electric Conductivity
Electric Impedance
Feasibility Studies
Humans
Image Processing, Computer-Assisted
Magnetic Resonance Imaging
Phantoms, Imaging
Tomography



View this publication's entry at the Pubmed website PMID: 22684125
jenkins-FCD Prod-158 788e756cee95f97989f23472575fc754037544be