Magnetic Resonance (MR) safety evaluations provide important information to patients and physicians regarding the potential hazards of implanted medical devices in the MR environment. Radio Frequency (RF) induced heating is one of the important safety concerns and is evaluated using a standard test method, ASTM F2182.¹ This test method, however, does not account for physiological cooling mechanisms such as blood flow, and as a result may produce overly conservative temperature measurements. These conservative test results are reported in the labeling of medical devices and can ultimately cause patients to be unnecessarily excluded from MRI procedures. Therefore, we hypothesize that a superior approach to MR safety evaluation of medical devices is to accurately simulate in vivo conditions. In order to test our hypothesis we developed a finite element simulation of RF induced heating with computational fluid dynamics, and compared the simulation results to experimental measurements a standard test phantom that was retrofit with a flow channel.The RF induced heating experiments were conducted using a 3.0 T Siemens Tim Trio MR system operating at a frequency of approximately 128 MHz. The RF body coil transmitter and a turbo spin echo sequence were used with a flip angle of 130°, TR/TE = 269/11 ms, TA = 450 s and a console reported Specific Absorption Rate (SAR) of approximately 5.5 W/kg achieved by overriding the system SAR monitor. The phantom was filled with a gelled saline prepared according to the ASTM F2182 standard with 25 L of distilled water mixed with 1.32 g/L NaCl and 10 g/L polyacrylic acid (PAA). The electrical conductivity of the gel at room temperature was 0.47 ± 10% S/m. The electrical and thermal properties of the gel mimic average human tissue but the viscosity of the gel prevents convective heat transport. The phantom was retrofit with a flow channel using silicone tubing with an inner and outer diameter of 7.9 mm and 11 mm, respectively. A nitinol stent, 10 mm x 80 mm (diameter x length), was positioned on the exterior of a thin-walled silicone tube (less than 1 mm wall thickness) with an outer diameter of 10 mm and connected in-line with the silicone tubing. Silicone tubing was routed from the MR equipment room through an RF waveguide and connected to the inlet and outlet of the phantom flow channel. A peristaltic pump was used to circulate room-temperature water from a reservoir through the silicone tubing at controlled flow rates. Temperature measurements were recorded using a fiber optic probe system (Luxtron FOT Lab Kit) with four STF type probes that take measurements at a frequency of 1 Hz with an accuracy of ± 0.2°C within 20°C of the calibration temperature (37°C). A temperature probe was placed at the proximal and distal end of the stent, between the stent and the thin-walled silicone tubing. The remaining probes were located on the opposite side of the phantom and in the water reservoir. The phantom was filled with gel to a depth of 9 cm and positioned on the patient table of the MR system without the spine coil and landmarked along the mid-length of the stent and phantom. COMSOL Multiphysics was used to solve the sequentially coupled electromagnetics, fluid dynamics and transient heat transfer of the RF induced heating of the stent with varied volumetric flow rates.A summary of the experimental measurements and simulation results are shown in Figure 1. The experiment and simulation of the average maximum temperature rise of the stent with respect to the volumetric flow rate of the fluid are in good agreement. The temperature rise of the stent was over 10°C without flow, and was reduced by 50% with a flow rate of only 50 mL/min. The vascular stent used in this study is indicated for the iliac arteries, which have volumetric flow rates of up to 450 mL/min. The simulation of a volumetric flow rate of 450 mL/min resulted in a 4.0°C temperature rise of the stent, which is a 60% decrease in the temperature rise compared to the zero flow case. A previous study has also demonstrated the thermal significance of flow for RF induced heating of a coronary stent.²The computer simulation enables accurate calculations of the influence of blood flow on the RF induced temperature rise of a vascular stent during MRI. Blood flow has a significant impact on the thermal response of tissue and stents during RF induced heating.