Tamer S Ibrahim¹, and Gary Boerger^2
1University of Pittsburgh, Pittsburgh, PA, United States, ²University of Oklahoma

Electromagnetic simulations in MRI can take significant calculation (CPU) time. Not only this can dampen the progress of evaluating the performance of the coil, it renders these methods impracticable for real time applications such as multi-transmission methods. This work aims at overcoming this obstacle through developing a stable and fast implicit finite difference time domain scheme that is suitable for MRI RF calculations. Very good correlation was attained for the B1+ field distribution obtained through calculations using standard FDTD and implicit FDTD methods, and through MRI experimental measurements.

Shumin Wang¹, Jeff Duyn², and Alan Koretsky^2
1NIH, Bethesda, Maryland, United States, ²NIH

Radio-frequency transmit coils need to produce homogeneous transverse magnetic fields. At low field strengths, this is typically accomplished by birdcage-like transmitters. At high field strengths, it is believed that due to destructive wave interferences, producing homogeneous fields by a single resonator is no longer possible. Here we present a general full-wave electro-dynamic approach that allows one to find the desired current distribution on any surfaces that generate homogeneous filed distributions on any specified locations inside a human-shaped phantom.

Paul R. Harvey¹, and Johan S. van den Brink^1
1Philips Healthcare, Best, Netherlands

Electromagnetic field simulations have been used to evaluate the effect on SAR and RF power demand during normal RF exposure, using the system body coil, for a human body model partially wrapped with a carbon (graphite) sleeve intended to shield parts of the body against the applied RF field. The presence of an RF absorbing/shielding sleeve creates an additional load inside the body coil which causes re-distribution of the E-field and requires increased input power together resulting in elevated global and local SAR.

Yong Pang¹, Daniel Vigneron^1,2, and Xiaoliang Zhang^1,2
1Radiology & Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States, ²UCSF/UC Berkeley Joint Graduate Group in Bioengineering, San Francisco & Berkeley, CA, United States

Radiofrequency penetration is not only determined by the coil itself, but also related to the electromagnetic properties (e.g. permittivity, conductivity) and geometry of imaging samples. In this work, the RF field penetration of surface coils in human liver and brain imaging at 7T is studied numerically. Due to conductivity effects in human liver and its irregular shape, RF penetration decreases while in the brain RF penetration increases due to “dielectric resonance” effect. Therefore, it is necessary to have larger size coils for liver than for brain to achieve the required penetration/coverage in ultrahigh field MRI.

Sukhoon Oh¹, Yeun Chul Ryu¹, Andrew Webb², and Christopher M Collins^1
1Radiology, College of Medicine, The Pennsylvania State University, Hershey, PA, United States, ²Radiology, The Leiden University Medical Center, Netherlands

In this study, we report recent progress towards a validation of in-vivo temperature change (T) of a human forearm, since in-vivo T is greatly moderated by perfusion. We acquired MR T map, then built a FDTD model from anatomical images to simulate SAR. Finally, we calculate T from the simulated SAR to validate the experimental T. In the T calculation, we considered the perfusion rate as a function of temperature, temperature decrease during acquisition of the second gradient-echo image, and the tuning condition of the RF heating coil. Qualitatively, the simulated T distribution appears very similar to the experimental one.

Xin Chen¹, and Michael Steckner^1
1Toshiba Medical Research Institute USA, Inc., Mayfield Village, OH, United States

While SAR is crucial for the RF safety control of an MR scan, B1rms (root mean square of the total B1 field ) serves as a supplemental safety metric. For example, B1rms is useful for defining the MR conditional labeling of implants. An unloaded quadrature driven (QD) birdcage coil generates nearly perfect B1 with circular polarization: B1+ (the tipping component) is homogeneous and B1- (the component rotating counter to the spin precession) is zero and thus |B1-|<<|B1+| is usually assumed. Loading the coil distorts the B1 field, causing B1+ and B1- components. While only B1+ is useful for MRI purposes, both components contribute to RF power deposition. In this work we use FDTD (Finite Difference Time Domain) numerical simulations to demonstrate that B1- should not be ignored in loaded coils at either 1.5T or 3T.

Andre Kuehne¹, Helmar Waiczies^1,2, Sairamesh Raghuraman³, Tobias Wichmann⁴, Titus Lanz⁴, Frank Seifert¹, and Bernd Ittermann^1
1Physikalisch-Technische Bundesanstalt, Berlin, Germany, ²Experimental and Clinical Research Center (ECRC), Max-Delbrueck Center for Molecular Medicine, Berlin, Germany, ³MRB Research Centre, Würzburg, Rimpar, Germany, ⁴Rapid Biomed, Rimpar, Germany

FDTD simulations are an important tool to calculate SAR values required for safe coil operation limits. After being presented with discrepancies between simulations using different software packages, the coil in question, a ³¹P birdcage coil, was simulated using three different FDTD programs (XFdtd, MWS, SEMCAD) and the results compared throughout the simulation volume. Possible sources of error, mainly different standards for scaling to different input power measures, are identified. When identical scaling standards are applied, the obtained results match very closely.

Zhangwei Wang¹, Owen Arthurs², Desmond T.B. Yeo³, and Fraser Robb^1
1GE Healthcare Coils, Aurora, OH, United States, ²2University of Cambridge, Cambridgeshire, United Kingdom, ³GE Global Research, Niskayuna, NY, United States

Pediatric body MR imaging is limited by the lack of dedicated coils. Many infants are typically imaged using adult head or knee coils, but the SAR consequences are unknown. Compared to adults, infants have several unique physiological and physical characteristics that may influence the thermal risk during RF exposure. In this work, we use EM numerical modeling to evaluate the local SAR deposition in an infant body model irradiated imaged by an adult head coil. SAR comparisons are performed for the infant model with infant-specific versus adult-specific electrical properties, and placed at different landmark positions.

Sebastian Wolf¹, and Oliver Speck^1
1Dept. Biomedical Magnetic Resonance, Otto-von-Guericke University, Magdeburg, Germany

In this work, deviations in the SAR prediction caused by simplification of the model are compared to deviations between highly detailed models, which do or do not reflect the actual patient. It will be shown, that model simplifications such as reduced extension and reduction of tissue types that may allow the creation of human models, which are more similar to the actual subject are feasible.

Sukhoon Oh¹, Giuseppe Carluccio², and Christopher M Collins^1
1Radiology, College of Medicine, The Pennsylvania State University, Hershey, PA, United States, ²Department of Electrical and Computer, University of Illinois at Chicago, IL, United States

According to prescription (such as 10-gram SAR for comparison to IEC limits), it possibly results an artificially high N-gram SAR region near air interfaces. Here, we implement a simple method for meaningful N-gram SAR calculation and provide a GUI for performing such calculations given an unaveraged SAR distribution. The tool was developed as a companion to a versatile MRI simulator (PSUdo MRI) which we recently released for public use, but also can be used with any unaveraged SAR dataset given in the right format. In comparison to the conventional 10-gram SAR calculation, our result shows smooth and actual 10-gram SAR.

Bu S Park^1,2, J McGarrity², Z Cao², K Sung³, S Oh², and C M Collins^2
1NIH, Bethesda, MD, United States, ²Radiology, The Pennsylvania State University, Hershey, PA, United States, ³Radiology, Standford University, Stanford, CA, United States

Except for RF shimming, transmit array pulses are generally expected to require more time than their quadrature-drive, single-pulse counterparts. Although many transmit array pulses can achieve better excitation uniformity than a simple pulse (including RF shimming), to do so in the same time duration with no significant increase in SAR is a challenge. Here we explore the effect of pulse duration in an array-optimized composite pulse (ACP) and in RF shimming designed to both improve excitation uniformity in the whole brain and reduce SAR using an 8-channel transmit head array, and compare results with the conventional quadrature drive at 7T. Minimum RF excitation time having better excitation uniformity and less SAR than the conventional quadrature drive for each pulse is presented.

Ulrich Katscher¹, Kay Nehrke¹, Peter Vernickel¹, Ingmar Graesslin¹, and Peter Börnert^1
1Philips Research Europe, Hamburg, Germany

RF shimming is able to compensate wave propagation effects at high main fields. The targeted B1 homogeneity has to be counterbalanced by the corresponding mean forward RF power required, which can be described by "regularized" RF shimming. However, this procedure frequently results in heterogeneous power distributions over the different transmit channels, which is disadvantageous for multi-channel RF amplifier design. To achieve comparable power requirements for all transmit channels, this study investigates "double-regularized" RF shimming, minimizing the mean as well as the maximum RF power. The approach was tested successfully in vivo for a whole-body, 8-channel TX/RX system at 3T.

Ozlem Ipek¹, Alexander J.E. Raaijmakers¹, Dennis W.J. Klomp², Johannes m Hoogduin², Peter R Luijten², Jan J.W. Lagendijk¹, and Cornelis A.T. van den Berg^1
1Radiotherapy, UMC Utrecht, Utrecht, Utrecht, Netherlands, ²Radiology, UMC Utrecht, Utrecht, Utrecht, Netherlands

The discrepancy if the B1+ value at prostate in measurement (10 µT ) and simulation (20 µT) is studied for 8 elements radiative antenna transmit surface array. The phase shimming was determined to obtain constructive B1+ interference in the prostate from the voltage phases of individual simulations. As the phase of the transmitted B1+ field of an element is not proportional to the feeding voltage but to the total current in an element we determined in simultaneously-driven elements, the voltage sources settings that would result in current phase setting from individual-driven elements. This solved the discrepancy problem.

Xiaoliang Zhang^1,2, Chunsheng Wang¹, Sarah Nelson^1,2, and Daniel Vigneron^1,2
1Dept of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, California, United States, ²UCSF/UC Berkeley Joint Graduate Group in Bioengineering, San Francisco and Berkeley, California, United States

In this work, we investigate the feasibility and advantages of ultrahigh high field body transmit array design using non-resonance method. With their features of uniform current and field distributions and frequency-insensitive, the proposed technique can be used for proton and multinuclear excitations with one single coil, providing a promising method to large-sized body coil designs at ultrahigh fields.

Wouter M Teeuwisse¹, Chris M Collins², Nadine B Smith¹, and Andrew G Webb^1
1Radiology, Leiden University Medical Center, Leiden, Netherlands, ²Radiology, Hershey Medical College

This work presents the results of electromagnetic simulations to study the effects of different high dielectric materials, size, geometry and position on the RF field distribution within the head at 7 Tesla. Local increases in signal-to-noise from materials with high dielectric constant are balanced by much smaller global decreases further away from the material. Splitting up the dielectric into several smaller subunits reduces the very strong wavelike behaviour that can occur when using materials with very high dielectric constants.

Matthias Korn¹, Simon Lambert¹, Xavier Maître¹, and Luc Darrasse^1
1IR4M (UMR8081), Université Paris-Sud XI - CNRS, Orsay, France

An efficient procedure for modeling and simulating phased-array coils is presented. The method uses a full wave EM simulation to calculate the EM-fields, while the effect of the port circuitry is substituted by the impedance of a discrete impedance. The value of this impedance is derived directly from the reflection coefficient of the preamplifier input and the coil impedance. The presented approach also includes sample losses, distributed capacitors, and geometrical and capacitive element decoupling. The method is demonstrated on a 6-channel receive-only coil to compare baseline sensitivity and g-factor performance of three different geometrical element configurations (gap, shared-conductor and overlap design).

Tyler Charlton¹, Adam Maunder¹, B. Gino Fallone^1,2, and Nicola De Zanche^1,2
1Dept. of Oncology, University of Alberta, Edmonton, Alberta, Canada, ²Dept. of Medical Physics, Cross Cancer Institute, Edmonton, Alberta, Canada

Three 8-channel array designs are compared at 3T: a) a gapped array; b) a standard overlapped array with zero mutual inductance between nearest neighbors; and c) a gapped array with mutual-inductance-nulling transformers between nearest neighbors. Preamp decoupling was used for all three. The gapped array provided the best SNR performance (respectively 18% and 24% better) since it had the lowest noise covariance.

Riccardo Lattanzi^1,2, and Daniel K Sodickson^1,2
1Center for Biomedical Imaging, New York University Langone Medical Center, New York, NY, United States, ²Radiology, New York University Langone Medical Center, New York, NY, United States

Using a rigorous full-wave electrodynamic formulation based on dyadic Green’s functions (DGF), we calculate ideal current patterns that result in the highest possible signal-to-noise ratio (SNR) in homogeneous spherical and cylindrical samples, and compare the resulting patterns to familiar coil designs. Our results demonstrate the natural emergence of quadrature birdcage-like reception current patterns for imaging near the center of the object, as opposed to distributed loop-butterfly surface quadrature patterns for locations closer to the surface. Ideal current patterns become more complex at high field, indicating that innovative coil designs may be needed in order to approach the optimal performance.

Scott B. King¹, Mike J Smith¹, and Boguslaw Tomanek^2
1Institute for Biodiagnostics, National Research Council of Canada, Winnipeg, Manitoba, Canada, ²Institute for Biodiagnostics (West), National Research Council of Canada, Calgary, Alberta, Canada

The number of receive channels on clinical MRI systems is now as high as 128, allowing the possibility for highly accelerated parallel MRI and increased SNR. To lessen the demands for data handling and processing, data compression methods are being explored. Here eigenmode channel compression was investigated for complex volume array designs revealing that significant (N/2) channel compression is possible while retaining >95% SNR and showing that symmetric, uniformly distributed array elements, as in the 24-channel head array, leads to more degeneracy and quad/antiqued pairs, and ultimately, whole-volume channel compression is possible, making permanent hardware channel compression an option.

Jörg Felder¹, and Nadim Joni Shah^1,2
1Institute of Neuroscience and Medicine-4, Forschungszentrum Juelich GmbH, Juelich, NRW, Germany, ²Department of Neurology, Faculty of Medicine, JARA, RWTH Aachen University, Aachen, Germany

Ultimate intrinsic SNR has been introduced to establish an upper limit for physically achievable SNR in MRI. Up to now, due to computational complexity, ultimate intrinsic SNR solutions have only been presented for highly symmetric and homogeneous phantoms. Using 3D field simulations, we have evaluated ultimate intrinsic SNR for the human head at 9.4T allowing comparison of novel conductor arrangements for high field, parallel imaging coils and evaluating experimental performance in an absolute measure. Simulation results using a human head model (Ella) are presented.

Alexander J.E. Raaijmakers¹, Cornelis A.T. van den Berg¹, and Dennis W.J. Klomp^2
1Radiotherapy, UMC Utrecht, Utrecht, Netherlands, ²Radiology, UMC Utrecht, Utrecht, Netherlands

Body imaging at 7 Tesla is hampered by the increased B1 signal attenuation at higher frequencies. The question therefore arises whether the potential gain in SNR by moving towards higher magnetic field strengths is diminished by the reduced sensitivity of receive coils. In this study, we compare potential receive array element designs and how their sensitivity patterns depend on the B1 frequency. It is shown that when the imaging target is located in the far-field of the antenna, the coil sensitivity does not decrease with magnetic field strength. This results in a quadratic increase in SNR with magnetic field strength.

Yong Pang¹, Bing Wu¹, Daniel Vigneron^1,2, and Xiaoliang Zhang^1,2
1Radiology & Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States, ²UCSF/UC Berkeley Joint Graduate Group in Bioengineering, San Francisco & Berkeley, CA, United States

The rapid development of parallel imaging requires high performance RF coil arrays with excellent decoupling performance. In this study, parallel imaging performance for a new 8-channel volume transceiver array with tilted microstrip elements was investigated in terms of reconstructed image quality, noise correlation matrix and g-factor. In vivo human knee images were obtained using a 7T whole body MR scanner. Commonly used parallel imaging methods – SENSE and GRAPPA – were utilized to perform accelerated image reconstructions. The results demonstrate excellent parallel imaging performance for our proposed tilted microstrip array.

Narayanan Krishnamurthy¹, and Tamer S Ibrahim^1
1University of Pittsburgh, Pittsburgh, PA, United States

A receive coil can distort the transmit field and cause local increased SAR. We seek to investigate the effects of a detuned and thin 32- loop receive-only head insert on transmit coil performance at 7 Tesla.

Hongbae Jeong¹, Suk-Min Hong¹, Joshua Haekyun Park¹, Myung-Kyun Woo¹, Young-Bo Kim¹, and Zang-Hee Cho^1
1Neuroscience Research Institute, Gachon University of Medicine and Science, Incheon, Korea, Republic of

This study is designed to compare the channel numbers of multi channel T/Rx coils performace at 7.0 Tesla. 2,4,6 and 8 channel multi T/Rx coils are constructed with same dimension on 270mm diameter of acryl case. The 6 channel coil showed the best SNR in central and peripheral area compared to 4 and 8 channel coils. For parallel imaging techinique, we prefer to use the 8 channel coil which displays similar SNR with 6 channel coil, and the 6 channel coil may have benefit of some angiography studies which requires more contrast diference in central area. The 2 channel coil has no specific advantage in terms of both SNR and Tx power consumption.

Arne Reykowski¹, Charles Saylor¹, and G. Randy Duensing^1
1ACD, Invivo Corporation, Gainesville, FL, United States

Noise coupling theory suggests that preamp decoupling methods may simplify coil tuning but have no impact on combined SNR. Based on this theory, the authors conducted two experiments involving a set of coupled receive coils. In the first experiment, both coils were matched using conventional preamp decoupling. In the second experiment, both coils were matched at resonance. While the individual channel signals for both cases were quite different, the combined SNR was very similar; this indicates that the theory is valid.

Mohamed Aly Saad Aly¹, Nibardo Lopez¹, Daniel Weyers², Sarbast Rasheed¹, Eihab M Abdel-Rahman¹, and Arsen Hajian^2,3
1System Design Engineering, University of Waterloo, Waterloo, Ontario, Canada, ²Tornado Medical Systems, Waterloo, Ontario, Canada, ³System Design Engineering, University of Waterloo, Waterloo, ontario, Canada

The possibility of integrating single-wall carbon nanotubes (SWCNT) in RF receivers was investigated in this work. We have approached this by two methods; replacing copper in copper coils by SWCNT and coating copper coils with sufficient thickness layer of SWCNT. Replacing copper with Single-wall carbon nanotubes shows promising results with some challenges. These hurdles are; high DC resistance and low impedance values. Using different CNT grades might help to overcome these challenges. Because of the high DC resistance of the CNT layer, coating copper with SWCNT layer did not affect the coil performance in terms of resistance, impedance, or quality factor.

Johanna Jacoba Bluemink¹, Anna Andreychenko¹, Astrid L.H.M.W. van Lier¹, Marielle Phillippens¹, Jan J.W. Lagendijk¹, Peter R. Luijten², and Cornelis A.T. van den Berg^1
1Radiotherapy, University Medical Center Utrecht, Utrecht, Netherlands, ²Radiology, University Medical Center Utrecht, Utrecht, Netherlands

The head and neck (H&N) region is challenging to image due to air cavities that distort the B0 field. At 7T better a SNR can be obtained, but susceptibility effects are increased and for the H&N region no commercial 7T transmit coils are available. Two setups, a volume T/R coil placed near the neck and a traveling wave setup are compared in transmit performance. With the volume T/R coil sufficient B1+ can be obtained for SE in the nasopharynx and the oropharynx. The traveling wave setup can cover a larger volume, but the maximum B1+ obtained was only 3ìT.

Matteo Pavan¹, Roger Lüchinger¹, and Klaas Paul Pruessmann^1
1Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland

Noise reduction is an important issue for SNR (Signal to Noise Ratio) maximization in MR measurements. Automatic Matching Networks (AMN) have been introduced and are a great tool to overcome notoriously difficult matching situation such as coil arrays, mechanically adjustable coils and in general coils that see different loading conditions. In this work a new automatic matching network has been implemented and tested, it allows automatic SNR measurements over thousands points in a short time, in a repeatable and robust way. An initial study of the behavior of the SNR has been done.

Mohan Lal Jayatilake^1,2, Judd Storrs^1,3, Wen-Jang Chu^1,3, and Jing-Huei Lee^1,4
1Center for Imaging Research, University of Cincinnati, Cincinnati, OH, United States, ²Department of Physics, University of Cincinnati, Cincinnati, OH, United States,³Department of Psychiatry and Behavioural Neuroscience, University of Cincinnati, Cincinnati, OH, United States, ⁴School of Energy, Environmental, Biological, and Medical Engineering, University of Cincinnati, Cincinnati, OH, United States

Optimal image quality for MRI at high field requires a homogeneous RF (B1) field; however, dielectric properties of the human brain cause B1 field inhomogeneities in the periphery of the head. Selecting the appropriate permittivity and the quantity of material for B1 shim is essential. In this work we introduce a theoretical framework for determining the requisite dielectric constant of the passive shim material.

Guillaume Ferrand¹, Michel Luong¹, Martijn A Cloos^1,2, Alain France¹, Alexis Amadon², Nicolas Boulant², and Luc Darrasse^3
1IRFU/SACM, CEA-Saclay, Gif s/ Yvette, France, ²I2BM/Neurospin, CEA-Saclay, Gif s/ Yvette, France, ³IR4M (UMR8081), Univ Paris-Sud, CNRS, Orsay, France

In high field MRI, a phased array coil comprising a large number of transmit elements appears more attractive with respect to local SAR reduction and to B1 inhomogeneity mitigation using parallel transmission. However, if each element were driven by a dedicated channel consisting of a concatenation of signal modulator and high power amplifier, the system cost would increase linearly with the number of transmit elements. It would be the same for the acquisition time of B1 maps necessary for parallel transmission. Here, we propose a new hardware scheme based on the singular value decomposition (SVD) to drive N elements with M channels, where M  N. The proof of concept is provided by numerical simulations of the whole process.

Alessandro Sbrizzi¹, Hans Hoogduin¹, Gerard L.G. Sleijpen², Astrid L Van Lier³, Jan J. Lagendijk¹, Peter Luijten¹, and Cornelis A.T. van den Berg^1
1Imaging Division, UMC Utrecht, Utrecht, Netherlands, ²Department of Mathematics, Utrecht University, Utrecht, Netherlands, ³UMC Utrecht

In this work we present a new approach for extracting complete complex information about the transmit and receive fields of a RF transmit coil from a B1+ amplitude measurement. The method is based on a Bessel/Fourier modes expansion of the B1+ and B1- maps. The results obtained from FDTD simulations and in vivo measurements confirm the validity of the novel approach.

Kawin Setsompop^1,2, and Lawrence L Wald^1,3
1Radiology, A. A. Martinos Center for Biomedical Imaging, MGH, Charlestown, MA, United States, ²Harvard Medical School, Boston, MA, United States, ³Harvard-MIT Division of Health Sciences and Technology, MIT, Cambridge, MA, United States

We propose a strategy which employs the “dark modes” of excitation coil array to cancel local E-fields that produce SAR hotspots. The method is based on the observation that some modes of an array coil (e.g. birdcage with the wrong circular polarization) produce E-fields but not useful excitation of MR magnetization (i.e. no B1+). We can therefore energize these “dark modes” exclusively with the goal of cancelling local SAR hot-spots. While this will not lower SARglobal (since more power will be used), our goal is local SAR reduction at a few hot spots where local SAR is the limiting factor.