Issues with radiofrequency heating in MRI

To the Editor: In recent years, the use of magnetic resonance imaging (MRI) has been increasing in patients with implanted devices, such as cardiac pacemakers and neurostimulators, which used to preclude the use of MRI owing to patient safety and device malfunction concerns.(1-6) In the past few years, however, more of these implant devices have become MR-compatible and can function properly during MRI studies. For these devices, one of the remaining limiting factors and safety considerations is radiofrequency (RF) heating, particularly in the conductive part of the implanted device which may damage the surrounding tissues. To allow the use of MRI with these types of devices, manufacturers typically specify the conditions under which imaging sequences can be safely used, including the maximum specific absorption rate (SAR) values and gradient strengths. For SAR values, clinical users often rely solely on the SAR values reported by the MRI system for specific MRI sequences. Therefore, the accuracy and consistency of the SAR values reported by the MRI system are become more relevant and critical for patient safety. However, to the best of our knowledge, these SAR values are not routinely verified or validated independently by clinical users anywhere in the today’s clinical practice. Although the physical principles of RF heating are simple and straightforward, accurate calculation of the SAR (measured in W/kg) is complicated by many factors, including patient size, heterogeneity of tissue conductivity, and differences in the RF power distribution profiles of the various MRI scanning sequences, as well as the specific scanning parameters. In general, SAR values increase with patient body weight. However, for the most part, the calculation of SAR values is proprietary for each MRI system manufacturer. Consequently, no independent validation or verification is available to clinical users, who must completely rely on the SAR value reported by the vendor. In a recent survey of the SAR values for clinical MRI spine studies at our institution, we found some inconsistencies and possible inaccuracies in the reported SAR values. As shown in Fig. 1, for the reported SAR values for the three-plane localizer scans at 1.5 T, some of the