Accurately measuring changes in body temperature is necessary during a magnetic resonance imaging (MRI) examination or during local hyperthermia treatments. In the case of an MRI examination, these variations, due to exposure to the radio frequency waves generated by the machine, are small, but regulations stipulate thresholds that must not be exceeded. Thus, throughout the duration of an MRI examination of the brain, the absolute temperature of the organ must remain below 39°C. 

But how can temperature variations be measured precisely and locally, no matter how small there are? Thanks to MRI! Currently, the thermometry method considered most accurate for small temperature variations is the so-called "proton resonance frequency shift" (PRFS): the proton resonance frequency is a parameter that can be measured by MRI and which has the good taste of varying with temperature. But other factors make this frequency vary: magnetic field drift, a phenomenon that is all the more important as the magnetic field is high, body movements related to breathing, heartbeats, etc...

A team from BAOBAB (NeuroSpin) proposes to improve the method of thermometry by taking into account the variations of the magnetic field in the calculation of the temperature.

The researchers have carried out a series of high-field thermometry experiments, with simultaneous monitoring of the magnetic field and local heating due to radiofrequencies. The experiments were carried out in vitro, with selected magnetic field perturbations, and in preclinical studies, on an anesthetized macaque monkey, at 50 and 100% of the maximum level of specific absorption rate (SAR, as for a cell phone) allowed by international regulations.

The accuracy of in vivo measurements was of the order of 0.15°C in areas little affected by movement. In the same region, the temperature increase reached 0.5 to 0.8°C after 20 minutes of heating at a 100% specific absorption rate and was about half as high at a SAR of 50%. 

With this first study conducted in vitro and in vivo, the authors conclude that the inclusion of magnetic field fluctuations in the image reconstruction is beneficial for the measurement of small temperature increases that occur during standard MRI brain examinations. Further work is needed to correct the field perturbations induced by subject movement and thus, ultimately, to accurately measure the temperature increases induced in any individual during an MRI examination.