Measuring local RF heating in MRI: Simulating perfusion in a perfusionless phantom

Purpose: To overcome conflicting methods of local RF heating measurements by proposing a simple technique for predicting in vivo temperature rise by using a gel phantom experiment. Materials and Methods: In vivo temperature measurements are difficult to conduct reproducibly; fluid phantoms introduce convection, and gel phantom lacks perfusion. In the proposed method the local temperature rise is measured in a gel phantom at a timepoint that the phantom temperature would be equal to the perfused body steady-state temperature value. The idea comes from the fact that the steady-state temperature rise in a perfused body is smaller than the steady-state temperature increase in a perfusionless phantom. Therefore, when measuring the temperature on a phantom there will be the timepoint that corresponds to the perfusion time constant of the body part. Results: The proposed method was tested with several phantom and in vivo experiments. Instead, an overall average of 30.8% error can be given as the amount of underestimation with the proposed method. This error is within the variability of in vivo experiments (45%). Conclusion: With the aid of this reliable temperature rise prediction the amount of power delivered by the scanner can be controlled, enabling safe MRI examinations of patients with implants. © 2007 Wiley-Liss, Inc.