Safety in MRI

Monday 1 June 2015

Frank Seifert¹, Gerd Weidemann¹, and Bernd Ittermann^1
1Physikalisch-Technische Bundesanstalt (PTB), Braunschweig und Berlin, Germany

The hazard of excessive local tissue heating due to metallic implants can be mitigated when utilizing the additional degrees of freedom of a transmit coil array. In this proof of principle study we used the single Q-matrix associated with the proximate tissue volume of a metallic implant to determine the 'worst case' steering conditions for implant heating. From a simple projection of the voltage vector to be applied to a sub space which is orthogonal to this 'worst case' voltage vector new steering conditions are determined which result in much lower implant heating. To demonstrate the efficacy of this method we performed thermal simulations using a 7T 8-channel head coil model. This approach can simply be embedded into B1-shimming procedures.

Thomas M Fiedler¹, Mark E Ladd^1,2, and Andreas K Bitz^1
1Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany, ²Erwin L. Hahn Institute for Magnetic Resonance Imaging, University Duisburg-Essen, Essen, Germany

It has been shown that specific configurations of high-permittivity pads can lead to an increase of local SAR. The electric field distribution depends on several parameters, e.g. polarization and distribution of the incident field, tissue distribution, and the actual geometry and position of the pad as well as its permittivity. In this work, we investigate several combinations of these parameters by RF field simulations in two body models and with two excitation scenarios to determine under which conditions SAR elevations are likely to occur or can be avoided.

Angel Torrado-Carvajal^1,2, Esra A. Turk^2,3, Joaquin L. Herraiz^2,3, Yigitcan Eryaman^2,4, Juan A. Hernandez-Tamames^1,2, Elfar Adalsteinsson^5,6, Larry L. Wald^4,6, and Norberto Malpica^1,2
1Medical Image Analysis and Biometry Lab, Universidad Rey Juan Carlos, Mostoles, Madrid, Spain, ²Madrid-MIT M+Vision Consortium, Madrid, Spain, ³Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge, MA, United States, ⁴Martinos Center for Biomedical Imaging, Dept. of Radiology, MGH, Charlestown, MA, United States, ⁵Dept. of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, United States, ⁶Harvard-MIT Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States

In this work we compare the results of B1+ field and SAR distribution obtained by using patient-specific 3PD images and two label fusion estimation approaches over a T1-weighted volume for fat and water segmentation. B1+ field distributions were found to be almost the same for the three models. An IDEAL label-fusion approach provides very similar SAR distribution results to the patient-specific approach. A CT label-fusion approach provides an increased SAR distribution map. The use of label fusion techniques to estimate the fat and water separation in MRI images allows an accurate segmentation with a similar accuracy as patient-specific 3PD sequences.

Matthew Restivo¹, C.A.T van den Berg¹, Alexander Raaijmakers¹, Peter Luijten¹, and Hans Hoogduin^1
1University Medical Center Utrecht, Utrecht, Netherlands

Parallel RF Transmit (pTx) in high field MRI is limited by the difficulty in correctly estimating the local specific absorption rate (SAR) given the complicated interferences of the electric field distributions produced by the various elements. Directional couplers allow for current to be measured easily on the transmission line, but parallel matching circuitry transforms the phase and amplitude of the current before entering the element. We propose a method for converting directional coupler values to tx element current values, thus allowing for the correct scaling and linear combination of simulated fields. This is necessary in order to accurately predict local SAR without overly complex simulations.

Manuel Murbach¹, Esra Neufeld¹, Eugenia Cabot¹, Earl Zastrow¹, Juan Corcoles², Wolfgang Kainz³, and Niels Kuster^1,4
1ITIS Foundation, Zurich, Switzerland, ²Department of Electronic and Communication Technology, Universidad Autónoma de Madrid (UAM), Madrid, Spain,³Center for Devices and Radiological Health (CDRH), US Food and Drug Administration (FDA), Silver Spring, MD, United States, ⁴Swiss Federal Institute of Technology (ETH), Zurich, Switzerland

In this study, we investigate the effects of pTx multitransmit body-coils at 3 T on anatomical models of a pregnant woman with her fetus at three different stages of gestation. Various imaging positions, regions of interest (ROI), and pTx coil excitation vectors are evaluated for effect on SAR and B1+ uniformity. The B1+ uniformity can increase by a factor of >4. The more equalized field distribution for large ROI typically leads to lower local SAR. However, the worstcase excitations identified lead to extremely high local SAR enhancement, emphasizing the need for robust management of exposure safety.

Mihir Pendse¹ and Brian Rutt^1
1Radiology, Stanford University, Stanford, CA, United States

Patient local SAR is usually controlled by way of estimations from numerical simulations on a library of tissue models. However, due to the variability between subjects in both anatomy shape and positioning within coil, mismatch between patient and model must be taken into account. We demonstrate that by using multiple body models with moderate degrees of mismatch it is possible to achieve significantly lower SAR than with a single model. The implication of this result is that, despite significant intersubject variability, relatively sparse body model libraries can be used without sacrificing robustness of SAR estimation.

Berk Silemek^1,2, Volkan Acikel^1,2, and Ergin Atalar^1,2
1Bilkent University, Ankara, Turkey, ²Umram, Ankara, Turkey

The aim of this study is to present a prototype of the implantable temperature sensor which will be used in animal studies to emulate the situation of a patient with an Active Implantable Medical Device inside the MRI and investigate possible tissue damage due to the RF induced tissue heating. The prototype records temperature information during MRI scan and transmits the desired data wirelessly, while implant is still inside of the body, to the remote without any surgery operation.

Matthew Restivo¹, Ronald Mooiweer¹, C.A.T van den Berg¹, Alexander Raaijmakers¹, Frank Simonis¹, Peter Luijten¹, and Hans Hoogduin^1
1University Medical Center Utrecht, Utrecht, Netherlands

We built a phantom which exhibits significant localized heating and provides good conditions for measuring temperature rise with MR thermometry and/or using fiber optic temperature probes. Measurements with this phantom allow for local SAR validation for RF coil models, a necessary step prior to relying on SAR simulations for in-vivo safety decisions.

Giuseppe Carluccio^1,2, Cem Murat Deniz^1,2, and Christopher Michael Collins^1,2
1Radiology, Center for Advanced Imaging Innovation and Research (CAI2R), New York University School of Medicine, New York, New York, United States,²Radiology, Bernard and Irene Schwartz Center for Biomedical Imaging, New York University, New York, New York, United States

Safety guidelines are provided by IEC to set maximum temperature and maximum SAR achievable during MRI scans. SAR-based safety evaluation of multichannel transmit systems has been accelerated using Virtual Observation Points to enable rapid RF safety monitoring and RF pulse design based on SAR constraints. In this work, using a matrix formalism to compute temperature, we propose a method to identify temperature based VOPs. Temperature based VOPs enable fast parallel transmit RF pulse design with iterative or non-iterative methods based on strict temperature constraints.

Vahid Ghodrati¹ and Abbas Nasiraei Moghaddam^1,2
1BME, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Tehran, Iran, ²School of Cognitive Sciences, Institute for Research in Fundamental Sciences (IPM), Tehran, Iran

Ferromagnetic objects in the MR-environment experienced linear magnetic region or saturation region. almost ferromagnetic objects placed in the saturation region in the commercial MR-scanners. Therefore, it is necessary to considered saturation region for these objects to have a better estimation of the torque and force that acted on ferromagnetic objects. Here, the approximated strength for the magnetic field that can be saturated the almost ferromagnetic objects will be presented. Also, it will show that just by using the linear region, the approximation of force will be 1.5 times overestimated in comparison by considering of real states for these objects.

Yujuan Zhao¹, Tiejun Zhao², and Tamer Ibrahim^1
1University of Pittsburgh, Pittsburgh, Pennsylvania, United States, ²Siemens Medical Solutions USA, Pittsburgh, Pennsylvania, United States

PTX excitation may create distinct hot spots inside the human body by constructive interference of electric fields generated by the respective coil elements driven with individual amplitudes and phases. Monitoring global and local peak SAR is a challenging task. In this study, the system global power and the local SAR from Virtual Observation Points (VOP) model in conjunction with rigorous RF modeling that incorporates coupling are demonstrated and verified in-vivo by experiments acquired using the Parallel RF transmission system. The input simulated SAR model is also verified quantitatively using B1+ maps as well as local VOP SAR monitoring.

Nanda Deepa Thimmappa¹, Christina Y Ahn², Silvina P Dutruel¹, Joshua L Levine³, Srikanth Reddy Boddu¹, and Martin R Prince^1
1Radiology, Weill Cornell Medical College, New York, NY, United States, ²NY Langone Medical Center, Department of Plastic Surgery, New York, New York, United States, ³Department of Plastic Surgery, New York Eye and Ear Infirmary of Mount Sinai, New York, New York, United States

There are anecdotal reports of tissue expander damage or dislodgement during MRI and are currently labeled “MR Unsafe”. At our institution, abdomen-pelvis MRI/MRA is frequently requested in these patients to map perforator vessels supplying abdominal fat/skin to plan autologous breast reconstruction. Since these tissue expanders will be removed at the time of breast reconstruction, surgeons performing these operations feel that the benefit of pre-operative MRA outweighs the risk of any implant migration or local tissue damage potentially caused by MRI. In this study we retrospectively review breast tissue expander safety for 73 patients who underwent abdominal and pelvic MR for autologous breast reconstruction.

Mihir Pendse¹ and Brian Rutt^1
1Radiology, Stanford University, Stanford, CA, United States

While shorter electromagnetic wavelengths and resulting SAR hotspots are a source of concern in ultra-high field MRI, the same phenomenon can be desirable in the field of targeted RF hyperthermia. In this work, we describe the “maxSAR” pulse design problem and implement a nonconvex interior-point algorithm to solve it. We demonstrate in simulation that an 8-loop pTx array operating at 298 MHz (7T), can be used to intentionally produce localized hotspots at arbitrarily located target regions while still constraining the SAR at healthy voxels to be within the local SAR regulatory limit.

Huan Li¹, Yan An¹, Qinli Sun¹, Yanyan Li¹, Pan Cao¹, Miaomiao Wang¹, Jianxin Guo¹, and Jian Yang^1
1Department of Radiology, The First Affiliated Hospital of Medical College,Xi'an Jiaotong University, Xi'an, Shaanxi, China

Purpose Magnetic resonance imaging (MRI) as a diagnostic method have been widelyPurpose Magnetic resonance imaging (MRI) as a diagnostic method have been widely clinical application for over twenty years. A prominent disadvantage of MRI is the high level of noise produced during scanning process. The noise is hazardous for patients and operators. It may cause the temporary or permanent hearing threshold shift. This study is aimed to explore whether the noise of 3T MRI has an influence on normal adults hearing. Methods We performed head MRI to 29 healthy volunteers who had passed the Auditory Brainstem Response (ABR) hearing test . We performed the ABR audiometry within 24 hours before the MRI examination for the first time, and carry on the second audiometry within 20 minutes after the MRI scanning, Among all the 29 cases, 26 showed a hearing threshold, and we review their hearing in 25 days later. The results were applied Wilcoxon matched-pairs test, SPSS 18. Results There was significant difference of left and right ear hearing between before and 20 minutes after MRI (p<0.01) and no significant difference of both ear hearing between before MRI and the review in 25 days later. Conclusion The noise lasted close to 1 hour from 3T MRI scanning has a transient effect on adults hearing.clinical application for over twenty years. A prominent disadvantage of MRI is the high level of noise produced during scanning process. The noise is hazardous for patients and operators. It may cause temporary or permanent hearing threshold shift. This study is aimed to explore whether the noise of 3T MRI has an influence on adults hearing. Methods We performed head 3T MRI to those 29 cases of healthy volunteers who had passed the Auditory Brainstem Response (ABR) hearing test . We performed the ABR audiometry within 24 hours before the MRI examination for the first time, and carry on the second audiometry within 20 minutes after the MRI scanning, and we review their hearing 25 days later. The results were applied Wilcoxon matched-pairs test, SPSS 18. Results The left and right ear¡¯s hearing before and after MRI , the tests were taken to be significant at p<0.01. There have not significantly difference in both ear¡¯s hearing before MRI and the review. Conclusion The noise that 3T MRI produced has only a transient effect on adults hearing.

Yacine Noureddine^1,2, Oliver Kraff¹, Mark E. Ladd^1,3, Karsten Wrede⁴, Gregor Schaefers², and Andreas K. Bitz^3
1Erwin L. Hahn Institute for MRI, University Duisburg-Essen, Essen, NRW, Germany, ²MR:comp GmbH, MR Safety Testing Laboratory, Gelsenkirchen, NRW, Germany, ³Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, BW, Germany, ⁴Clinic for Neurosurgery, University Hospital Essen, Essen, NRW, Germany

Synopsis: For safety assessment of metallic aneurysm clips, RF-induced heating resulting from electric field elevations in the tissue is the major concern with respect to patient safety. In this study, RF and thermal simula-tions were conducted to determine RF heating around an aneurysm clip at 7T with a standard RF transmit coil using both a homogeneous and heterogeneous model of the head.

Joaquin L. Herraiz¹, Yigitcan Eryaman^1,2, Esra Abaci Turk¹, Angel Torrado-Carvajal^1,3, Adrian Martin^1,4, Emanuele Schiavi^1,4, Bastien Guerin⁵, Elfar Adalsteinsson^1,6, Lawrence L. Wald^5,7, Juan A. Hernandez-Tamames^1,3, and Norberto Malpica^1,3
1Madrid-MIT M+Vision Consortium in RLE, Massachusetts Institute of Technology, Cambridge, MA, United States, ²Center for Magnetic Resonance Research,Department of Radiology, University of Minnesota, Minneapolis, MN, United States, ³Medical Image Analysis and Biometry Laboratory, Universidad Rey Juan Carlos, Madrid, Spain, ⁴Dept. of Applied Mathematics, Universidad Rey Juan Carlos, Mostoles, Madrid, Spain, ⁵Martinos Center for Biomedical Imaging, Dept. of Radiology, MGH, Charlestown, MA, United States, ⁶Dept. of Electrical Engineering and Computer Science, Harvard-MIT Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States, ⁷Harvard-MIT Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States

Current approaches to acquire clinical 7-T MR data within acceptable SAR limits require imaging with reduced flip angle and/or number of slices, which reduces SNR/Contrast/Detectability, thereby undermining the potential benefits of high-resolution imaging. Parallel transmission RF pulse design has recently shown to be able to manage global and local SAR in 7-T systems if a correct patient head and torso model is known. Nevertheless, it is unclear if patient-specific SAR reduction methods could be implemented in a feasible way in the clinical practice in the near future. In this work, we propose an effective protocol for enabling patient-specific SAR modeling and reduction to utilize the full capability and clinical promise of 7-T MR imaging.

Esra Neufeld¹, Manuel Murbach¹, and Niels Kuster^1,2
1IT'IS Foundation for Research on Information Technologies in Society, Zurich, Switzerland, ²Swiss Federal Institute of Technology (ETHZ), Zurich, Switzerland

A fast method to estimate exposure safety during MRI based on thermal dose (CEM43) has been developed that is capable of considering the transient nature of heating, exposure duration, local SAR, and the impact of perfusion as well as thermoregulation. Thus it overcomes problems associated with traditional SAR-based limits. The assumptions and relationships underlying the model have been verified and necessary constants derived using simulations. Comprehensive uncertainty analysis has been performed and the predictions have been validated against measurement data and detailed simulations of realistic scan sequences and exposure scenarios.

Daisy Q Huang¹, Martin Prince¹, George Shih¹, and Yan Cao^1
1Radiology, New York Presbyterian Hospital/Weill Cornell, NY, New York, United States

This retrospective study explores the correlation between high T1 signal in dentate nuclei and multiple Gd contrast administrations. Signal changes vary with type of contrast agent used, suggesting a multifactorial basis beyond just number of Gd administrations.

Annie Papadaki^1,2, David Carmichael³, Mark James White^1,2, Hoskote Chandrashekar¹, Tarek Yousry^1,2, Beate Diehl^4,5, Louis Lemieux⁴, and John Stephen Thornton^1,2
1Lysholm Department of Neuroradiology, National Hospital for Neurology and Neurosurgery, UCLH, London, London, United Kingdom, ²Department of Brain Repair and Rehabilitation, UCL Institute of Neurology, London, London, United Kingdom, ³Imaging and Biophysics Unit, UCL Institute of Child Health, London, United Kingdom, ⁴Department of Clinical and Experimental Epilepsy, UCL Institute of Neurology, London, United Kingdom, ⁵Department of Neurophysiology, National Hospital for Neurology and Neurosurgery, UCLH, London, United Kingdom

In this study we assessed temperature changes (∆T) during MRI at the tip of Stereo-electroencephalography (SEEG) electrodes in a phantom. Fibre optic thermometry measurements at the tip of two electrodes were performed at 1.5T and 3T, during a high SAR TSE sequence and an fMRI EPI sequence. Different electrode tail terminations were tested to match possible clinical scenarios. Scanning with electrode tails only, the maximum ∆T did not exceed 1°C at 1.5T, while it exceeded the guideline limits at 3T. Connecting extension cables and varying the cable termination has a significant effect in RF-induced heating

Sachiko Yamaguchi-Sekino¹, Masaki Sekino², and Toshiharu Nakai^3
1National Institute of Occuaptional Safety and Health, Japan, Kawasaki, Kanagawa, Japan, ²Graduate School of Engineering, The University of Tokyo, Tokyo, Japan, ³Neuroimaging & Informatics, National Center for Geriatrics and Gerontology, Aichi, Japan

Occupational exposure to the static magnetic field (SMF) is an arising problem in operators of magnetic resonance imaging (MRI) system. Although the individual motion control is recommended for MR workers to prevent these temporal effects, more feasible safe working procedure is required to achieve the compliance monitoring. The present study set 30 cm of the approach restricted area from the end of 3 T MRI system instead of the individual motion control. The average of the maximum SMF exposure among subjects was decreased in 25 % by this action (t-test, p<0.01). And motion analysis by motion capture indicated that work performances were not affected under this action. These results suggest that this simple safe working procedure is applicative for MR workers to reduce occupational SMF exposure without remarkable changes in work performances. And the quantification of MR workers’ motions by motion capture provides the useful information for electromagnetic field dosimetry.

Bixia Chen^1,2, Tobias Schoemberg^1,2, Oliver Kraff¹, Andreas K. Bitz^1,3, Harald H. Quick^1,4, Mark Edward Ladd^1,3, Ulrich Sure², and Karsten Henning Wrede^1,2
1Erwin L. Hahn Institute for Magnetic Resonance Imaging, University Duisburg-Essen, Essen, NRW, Germany, ²Department of Neurosurgery, University Hospital Essen, University Duisburg-Essen, Essen, NRW, Germany, ³Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, BW, Germany, ⁴High Field and Hybrid MR Imaging, University Hospital Essen, University Duisburg-Essen, Essen, NRW, Germany

The importance of ultra-high-field MRI in neuroimaging is increasing rapidly. Titanium cranial fixation plates (CFP) are commonly used during neurosurgical operations, and have been considered MR-conditional in simulations and head model studies. We evaluated imaging artifacts in multiple sequences correlated to CFP in vivo. Five patients were examined before, within 72 hours after, and 3 months after surgery at 3T, and in a 7T whole-body MR system using an 8-channel RF head coil. CFP caused minor artifacts in TOF and MPRAGE at 7T, depiction of adjacent brain tissue was impaired in SWI due to susceptibility artifacts, comparable with 3T artifacts.

Thomas M Fiedler¹, Aaron S Kujawa¹, Frank Resmer², Patrick Stein², Titus Lanz², Mark E Ladd^1,3, and Andreas K Bitz^1
1Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany, ²RAPID Biomedical GmbH, Rimpar, Bavaria, Germany, ³Erwin L. Hahn Institute for Magnetic Resonance Imaging, University Duisburg-Essen, Essen, Germany

Numerical simulations for a bilateral 4-channel transmit/receive breast coil were performed to investigate the optimum excitation mode of the coil and to determine the maximum permissible input power. An exact model of the breast coil was generated and validated by a comparison of measured and simu-lated B1+ distributions in a phantom. An anatomical heterogeneous body model was used to obtain realistic specific absorption rates (SAR). B1+ shimming was performed to find an excitation mode with optimal B1+ homogeneity.

Simone Angela Winkler¹, Paul Picot², Michael Thornton², and Brian K Rutt^1
1Dept. of Radiology, Stanford University, Stanford, CA, United States, ²Endra Inc., Ann Arbor, Michigan, United States

We propose a new concept for direct measurement of specific absorption ratio (SAR), to be used as a safety assessment / monitoring tool for MRI. The concept uses short bursts of RF energy and measures the resulting thermoacoustic excitation pattern by an ultrasound transducer array, followed by image reconstruction to yield the 3D SAR distribution. We conducted an experimental study to test the feasibility of this novel concept for direct SAR mapping. The promising results encourage us to further develop the method for application to in-vivo local SAR mapping.

Xin Chen¹, Charles Poole², Michael Steckner¹, and Robert Brown^2
1MR, Toshiba Medical Research Institute USA, Inc., Mayfield Village, OH, United States, ²Department of Physics, Case Western Reserve University, Cleveland, OH, United States

Numerical simulations are used to investigate thermal impacts of 3.2W/kg head SAR limit for brain imaging. No significant SAR or temperature rise are observed in brain tissues or eyes.