The advent of parallel transmission at high field has led to many studies aiming at quantifying more accurately the Specific Absorption Rate (SAR)/temperature aspects while understanding in more details the risks involved in MRI experiments. This talk gives a review of different techniques employed for SAR calculations, their validations and real-time supervision. Safety margins arising from each step of the evaluation chain are described and future directions for temperature evaluation are presented.
Global and local SAR control in single-channel transmit systems
In single-channel transmit systems, a relatively conservative and experimental assessment of the global SAR can be performed. If forward and reflected powers indeed are measured, energy conservation dictates that the global SAR should not exceed the absorbed power divided by the mass of the subject. Peak 10-g SAR can then be deduced analytically for specific geometries1 or via numerical simulations. A safety k-factor can then be computed with k=max(SAR10g)/SARGlobal which thereby enables local SAR control via simple power monitoring. Alternatively, subject-based local SAR can be estimated with B1 map measurements and processing2,3, but it is not yet as mature as the widespread use of numerical simulations.
Global and local SAR control in multi-transmit channel systems
In pTx, the total electric field at a given location is the linear superposition of individual E-fields driven by N independent transmit channels. Yet again the measurement of the forward and reflected powers on each channel allows for a subject-based assessment of the global SAR by invoking energy conservation. With a quick calibration scan, global SAR thereby can be predicted for any arbitrary RF excitation4,5. Estimation of local SAR yet to date still necessitates the use of numerical simulations. Although experimental6 and computational7 efforts have been made towards the use of subject-based models for accuracy, the use of generic models combined with safety factors still prevails due to its simplicity of implementation. To take into account properly the coupling between coil elements, co-simulation should be used to reproduce the correct scattering matrix8. Alternatively, to achieve the same goal linear combinations of simulated basis field elements can be fitted to complex B1 measurements9,10 for known geometries. Resulting errors between simulated and measured transmit B1 fields can then be propagated to estimate 10-g SAR uncertainties11 or N2 individual MR thermometry experiments12 can be carried out to measure directly the so-called SAR Q matrices13. Another safety factor can be added to account for anatomic variability by simply simulating several models and positions14-17 and estimate the risks. Finally, additional safety factors can be enforced depending on the information and uncertainties returned by the monitoring hardware6. The number of Q-matrices, normally defined for each voxel, can be compressed to return Virtual Observation Points (VOPs)18 and drastically ease their incorporation in pulse design algorithms19,20 as well as real-time SAR supervision based on measured waveforms.
Temperature control in multi-transmit channel systems
Temperature, or even thermal dose21, is more relevant from the safety point of view than SAR. To what extent the local temperature rises however highly depends on the location of the SAR hotspots and the amount of perfusion. As shown by Massire and co-workers22, over thousands of random pTx RF pulses each saturating the IEC local SAR limits at 7T, in simulation very few posed problems in terms of temperature. It was also shown with some simulations performed at 10.5 T that the indirect link between temperature and SAR in fact could be exploited to gain further scan performance and increase safety23. At the cost of pre-simulations, the temperature calculation formalism was also shown to become strictly identical to the one used for SAR so that temperature VOPs likewise could be derived23. Great progress thus has been made from the numerical point of view to make temperature management in RF pulse design and exam supervision a reachable goal. Debates, however, still exist about the validity of the Pennes’ bioheat equation often used and alternate models have been proposed24,25. Experimental efforts have been made to characterize better the relationship between SAR and temperature25-29 but they remain limited in number due to the difficulty of performing non-invasive small temperature rise measurements in vivo. More work hence is needed to investigate bioheat models further. Nevertheless, it should be kept in mind that current SAR limits provided in the international guidelines, perhaps indeed of empirical value, are meaningful only if a thermal rationale supports them.
1. Hoult DI, Lauterbur PC. The sensitivity of the zeugmatographic experiment involving human samples. J Magn Reson 1979;34:425–433.
2. Voigt T et al. Patient-individual local SAR determination: in vivo measurements and numerical validation. Magn Reson Med 2012;68:1117-1126.
3. Katscher U, Findeklee C, Voigt T. B1-based specific energy absorption rate determination for nonquadrature RF excitation. Magn Reson Med 2012;68:1911-1918.
4. Zhu Y et al. System and SAR characterization in parallel RF transmission. Magn Reson Med 2012;67:1367-1378.
5. Deniz CM et al. Subject- and resource-specific monitoring and proactive management of parallel RF transmission. Magn Reson Med 2015;76:20-31.
6. Graesslin I et al. Comprehensive RF safety concept for parallel transmission systems. ISMRM 16 (2008) 74.
7. Villena JF et al. Fast electromagnetic analysis of transmit RF coils based on accelerated integral equation methods. ISMRM 22 (2014) 0623.
8. Kozlov M, Turner R. Fast MRI coil analysis based on 3-D electromagnetic and RF circuit co-simulation. J Magn Reson 2009;200:147-152.
9. Tiberi G et al. Evaluation of 3D radio-frequency electromagnetic fields for any matching and coupling conditions by the use of basis functions. J Magn Reson 2015;261:38-42.
10. Beqiri A et al. Comparison between simulated decoupling regimes for SAR prediction in parallel transmit MRI. Magn Reson Med 2015;74:1423-1434.
11. Ferrand G et al. Mathematical tools to define SAR margins for phased array coil in-vivo applications given E-field uncertainties. ISMRM 23 (2015) 1862.
12. Alon L et al. Method for in situ characterization of RF heating in parallel transmit MRI. Magn Reson Med 2013;69:1457-1465.
13. Graesslin I et al. A SAR prediction concept for parallel transmission MR. Magn Reson Med 2012;68:1664-1674.
14. de Greef M et al. SAR intersubject variability in 7T parallel transmit MRI of the head. Magn Reson Med 2013;69:1476-1485.
15. Shao Y, Zeng P, Wang S. Statistical simulation of SAR variability with geometric and tissue property changes by using the unscented transform. Magn Reson Med 2015;73:2357-2362.
16. Wolf S et al. SAR simulations for high-field MRI: how much detail, effort, and accuracy is needed? Magn Reson Med 2013;69:1157-1168.
17. Le Garrec M et al. Probabilistic analysis of the SAR intersubject variability safety factor in parallel transmission MRI. Magn Reson Med 2016;78:1217-1223.
18. Eichfelder G, Gebhardt M. Local SAR control for parallel transmission by virtual observation points. Magn Reson Med 2011;66:1468–1476.
19. Guerin B et al. Local SAR, global SAR, transmitter power, and excitation accuracy trade-offs in low flip-angle parallel transmit pulse design. Magn Reson Med 2014;71:1446–1457.
20. Hoyos-Idrobo A et al. On variant strategies to solve the magnitude least squares optimization problem in parallel transmission pulse design and under strict SAR and power constraints. IEEE Trans Med Imaging 2014;33:739–748.
21. Van Rhoon GC et al. Thermal dose thresholds: a potential guide for magnetic resonance radiofrequency exposure levels? Eur Rad 2013;23:2215-2227.
22. Massire A et al. Thermal simulations in the human head for high field MRI using parallel transmission. J Magn Reson Imag 2012;35:1312-1321.
23. Boulant N et al. Direct control of the temperature rise in parallel transmission by means of temperature virtual observation points: simulations at 10.5 Tesla. Magn Reson Med 2016;75:249-256.
24. Kotte A et al. A description of discrete vessel segments in thermal modelling of tissues. Phys Med Biol 1996;41:865-884.
25. Shrivastava D et al. RF heating in porcine models with a large 32 cm internal diameter, 7T (296 MHz) head coil. Magn Reson Med 2011;66:255-263.
26. Simonis FFJ et al. Feasibility of measuring thermoregulation during RF heating of the human calf muscle using MR based methods. Magn Reson Med 2016;75:1743-1751.
27. Boulant N et al. FID navigator-based MR-Thermometry method to monitor small temperature changes in the brain of ventilated animals. NMR in biomedicine 2015;28:101-107.
28. Nadobny J et al. Experimental investigation and histopathological identification of acute thermal damage in skeletal porcine muscle in relation to whole-body SAR, maximum temperature, and CEM43°C due to RF irradiation in an MR body coil of birdcage type at 123 MHz. Int J Hyperthermia 2015;31:409-420.
29. Oh S et al. Measurement of SAR-induced temperature increase in a phantom and in vivo with comparison to numerical simulation. Magn Reson Med 2014;71:1923-1931.