Magnetic susceptibility effects on the accuracy of MR temperature monitoring by the proton resonance frequency method

PURPOSE: To evaluate the error of MR temperature assessment based on the temperature-dependent Larmor frequency shift of water protons, which can result from susceptibility effects caused by the radiofrequency (RF) applicator. MATERIALS AND METHODS: Local frequency shifts due to RF applicator displacements were simulated numerically by means of a three-dimensional elementary dipole model. Experimental examinations using a water tank phantom equipped with a high-precision screw thread were applied to examine temperature and movement effects for five commercially available, MR-compatible RF applicators. Measurements were performed at 1.5 Tesla. RESULTS: For single-needle electrodes perpendicular to the external field, a distortion of 0.1 ppm and 0.2 ppm was recorded at a distance of 17.5 mm and 12.5 mm, respectively, to the needle shaft. Cluster applicators and umbrella-shaped applicators caused distortions of 0.1 ppm up to distances of 36 mm. Sinusoidal dependence on applicator orientation was found with the highest values for perpendicular orientation and the lowest values for orientation parallel to the magnetic field. With a single electrode oriented perpendicular to the field at a distance of 1.5 cm and 2.0 cm, a needle displacement of 5 mm led to an error in temperature measurement of 16.3 degrees C and 7.5 degrees C, respectively. CONCLUSION: In MR temperature measurement, displacement of the RF applicator by patient movement or breathing leads to significant errors that have to be taken into account when PRF temperature maps are used to monitor tumor ablation in the presence of paramagnetic applicators.