Temperature Dependence of the Susceptibility of Fat Leads to Significant Temperature Errors in PRFS-based MR Thermometry

摘要

Proton resonance frequency shift (PRFS)-based MR thermometry is hampered by temporal field changes. Temporal changes in the susceptibility distribution lead to field changes and are therefore a possible source of errors. The susceptibility of fat,chi_(fat) is temperature dependent, in the same order of magnitude as the temperature dependence of the chemical shift of water (0.01 ppm/degC). PRFS-based temperature measurements may therefore be corrupted by non-local field effects due to temperature induced susceptibility changes in fatty tissue. We performed simulations to quantify the influence of d chi_( fat)/dT on PRFS-based MR temperature maps during thermal interventions in the breast. Susceptibility distributions were calculated for a 3D breast model without and with a stationary Gaussian temperature distribution and a maximum temperature of 57degC. Subsequently, the magnetic field was calculated in Fourier-domain from these susceptibility distributions. Changes in the magnetic field distribution were quantified by subtraction of the pre- and post-heating outcome. Our simulations show that the temperature dependence of chi_(fat) leads to significant errors in PRFS-based MR temperature measurements in nearby glandular tissue during thermal interventions in the female breast. The maximum field change within the glandular tissue was 0.13 ppm, corresponding to a PRFS-based MR temperature error of DELTA T_(err)=13 degC. The Important to stress is that fat-suppression is not a solution for this effect.