TwinTree Insert

09-06 Temperature

he influence of temperature on relaxation times is well known from ana­ly­ti­cal NMR (Figure 09-14). Temperature also influences the diffusion coefficient and the chemical shift of the water peak.

Figure 09-14:
Drastic change of temperature. The white curve shows the decrease of signal in­ten­si­ty of a T1-weight­ed pulse sequence be­fore, during and after lo­cal heating (red curve) of brain tissue in an ex vivo ex­­pe­­ri­­ment. The temperature changes from 25° C to more than 60° C; relative SI drops by 50%.
SI = relative signal intensity; tem­pe­ra­tu­re in °Celsius, time (t) in minutes.

Thus, the question arose if in MR imaging temperature changes in the human body may influence relaxation times of tis­sues and therefore contrast. This might occur, for instance, in patients running high temperatures one day when undergoing MR and having normal tem­pe­ra­tu­res during a follow-up examination.

However, relaxometric measurements proved that any differences created are with­in the system error and are not to be expected to influence contrast in imag­ing of patients [⇒ Rinck 1987].

spaceholder redThermometry. The commonly used approach to mag­ne­tic re­so­nan­ce ther­mo­me­try is not based upon relaxation times mea­surements, but on changes of the re­­so­­nan­­ce fre­quen­cy caused by temperature changes.

In water, the electrons shield the nucleus from the magnetic field and thus de­crease the hydrogen resonance frequency. How­ever, as the temperature in­crea­ses, hy­dro­­gen bonds re­or­ga­ni­ze and the electron shield of the protons from the magnetic field gets even stronger, reducing the net field the protons are exposed to.

Their reso­nance frequency increases and this change can be measured and re­la­ted to temperature. This process is described as as proton reso­nance frequency shift (PRF or PRFS) ther­mometry. It is calculated from a series of gradient echo images [⇒ Rieke 2008]. Temperature-related effects can be mapped dynamically [⇒ Hy­ny­nen 1997, ⇒ Le Bihan 1989, ⇒ Matsumoto 1994].