Purpose: The increasing number of clinically oriented MRI studies at 7 T motivates the safety assessment of implants, since many 7 T research sites conservatively exclude all subjects with metallic implants, regardless of type or location. The purpose of this study was to investigate potential RF-induced heating during a 7 T MRI scan using a self-built transmit/receive RF coil in patients with implants used for refixation of the bone flap after craniotomy. Going beyond standard ASTM safety tests, a comprehensive test procedure for safety assessments at 7 T is presented which takes into account the more complex coupling of the electromagnetic field with the human body and the implant as well as polarization effects.
Methods: The safety assessment consisted of three main investigations using (1) numerical simulations in simplified models, (2) electric and magnetic field measurements and validation procedures in homogeneous phantoms, and (3) analysis of exposure scenarios in a heterogeneous human body model including thermal simulations. Finally, 7 T in vivo images show the degree of image artifact around the implants.
Results: The simulations showed that the field distortions remain localized within the direct vicinity of the implants. A parallel E-field polarization was found to be the most relevant component in creating local SAR deviations, resulting in a 10% increase in 10-g-averaged SAR and 53% in 1-g-averaged SAR. Using a heterogeneous human head model, the implants caused field distortions and SAR elevations in the numerical simulations which were distinctly lower than the maximum local SAR value caused by the RF coil alone. Also, the position of the maximum 10-g-averaged SAR remained unchanged by the presence of the implants. Similarly, the maximum absolute local temperature remained below 39 °C in the thermal simulations. Only minor artifacts from the implants were observed in the in vivo images that would not likely affect the diagnostic image quality in patients.
Conclusions: The findings suggested no evidence for noteworthy RF-related heating in humans after craniotomy using the described implants and for the particular RF coil that was used in this study. Here, identical transmit power restrictions apply with or without the implants. For other RF coils, the maximum permissible input power should be reduced by 10% until further simulations may indicate otherwise.