Finite element method‐based approach for radiofrequency magnetic resonance coil losses estimation. Issue 4 (31st January 2017)

• Record Type: Journal Article
• Authors: Giovannetti, Giulio; Tiberi, Gianluigi; Tosetti, Michela
• Is Part Of: Concepts in magnetic resonance. Volume 46B:Issue 4(2016)
• Page Start: 186
• Page End: 190
• Publication Date: 2017-01-31
• Subjects: numerical methods -- radiofrequency coils -- wire; Magnetic resonance -- Periodicals; Nuclear magnetic resonance -- Periodicals; Biomedical engineering -- Periodicals; Magnetic Resonance Spectroscopy -- Periodicals; 538.36
• DOI: 10.1002/cmr.b.21348
• Languages: English
• ISSNs: 1552-5031

Abstract

Abstract: The simulation and the design of radiofrequency (RF) coils are fundamental tasks to maximize Signal‐to‐Noise Ratio (SNR) in Magnetic Resonance (MR) applications. The estimation of coil resistance, that is, the losses within the coil conductors, which depends on tuning frequency, allows the prediction of coil performance and data SNR. At RF, the conductor resistance is increased due to the skin effect, which distributes the current primarily near the conductor surface instead of uniformly over the cross section. Moreover, the radiative losses estimation as a function of tuning frequency permits a total coil performance characterization, especially for high‐frequency tuned coils when this loss mechanism could be the dominant one. In this work we compared Finite Element Method (FEM) simulations with analytical calculations performed in wire loop RF coils for MR applications. Our results showed that FEM can predict the losses within the coil conductors at 5.7 MHz with a relative difference of <3% compared to analytical calculation, while the relative difference increased to 58% at 127.8 MHz. Concerning the radiative losses, the relative difference between analytical formulation and FEM was lower than 3% at 5.7 MHz, and increasing to 44% at 127.8 MHz. Experimental measurements on a circular coil prototype were also performed at 85.2 MHz and 127.8 MHz, showing a better agreement with FEM simulations than with analytical calculations.