The Proton-Resonance-Frequency-Shift Method Compared with Molecular Diffusion for Quantitative Measurement of Two-Dimensional Time-Dependent Temperature Distribution in a Phantom

Noninvasive thermometry methods with magnetic resonance imaging usually explore the temperature dependence of the molecular diffusion coefficient of water. A method based on the temperature dependence of the proton resonance frequency is proposed in this study and compared with the diffusion method. The comparison was made with a gel phantom with muscle characteristics and for a voxel size of 0.8 × 0.8 × 10 mm³. The root-mean-square deviation of the temperature images obtained with simulations of the thermal process is between 0.1 and 0.15°C for the proton-resonance-shift-based method with an acquisition time of 1 minute and 0.9-1°C for the diffusion-based method with an acquisition time of 4.5 minutes. Unfortunately, the proton-resonance-shift method is very sensitive to the drift of the external magnetic field and therefore a method of external references was proposed to correct for this drift. The method proves to be adequate as long as the thermal be of interest do not take more than 1 hour.