Temperature-induced tissue susceptibility changes lead to significant temperature errors in PRFS-based MR thermometry during thermal interventions.

Proton resonance frequency shift-based MR thermometry (MRT) is hampered by temporal magnetic field changes. Temporal changes in the magnetic susceptibility distribution lead to nonlocal field changes and are, therefore, a possible source of errors. The magnetic volume susceptibility of tissue is temperature dependent. For water-like tissues, this dependency is in the order of 0.002 ppm/°C. For fat, it is in the same order of magnitude as the temperature dependence of the proton electron screening constant of water (0.01 ppm/°C). For this reason, proton resonance frequency shift-based MR thermometry in fatty tissues, like the human breast, is expected to be prone to errors. We aimed to quantify the influence of the temperature dependence of the susceptibility on proton resonance frequency shift-based MR thermometry. Heating experiments were performed in a controlled phantom set-up to show the impact of temperature-induced susceptibility changes on actual proton resonance frequency shift-based temperature maps. To study the implications for a clinical case, simulations were performed in a 3D breast model. Temperature errors were quantified by computation of magnetic field changes in the glandular tissue, resulting from susceptibility changes in a thermally heated region. The results of the experiments and simulations showed that the temperature-induced susceptibility changes of water and fat lead to significant errors in proton resonance frequency shift-based MR thermometry.