The noninvasive thermometry method is based on the temperature dependence of the proton resonance frequency (PRF). High-quality temperature images can be obtained from phase information of standard gradient-echo sequences with an accuracy of 0.2 degrees C in phantoms. This work was focused on the in vivo capabilities of this method. An experimental setup was designed that allows a qualitative in vivo verification. The lower-leg muscles of a volunteer were cooled and afterwards reheated with an external water bolus. The temperature of the bolus water varied between 17 degrees C and 37 degrees C. The in vivo temperature images can be used to extract the temperature in muscle tissue. The data in the fat tissue are difficult to interpret because of the predominance of susceptibility effects. The results confirm the method's potential for hyperthermia control.

The temperature dependence of proton resonance frequency (PRF) is related to the temperature dependence of the screening constant and of the volume susceptibility constant. To evaluate the relative importance, an experimental setup was designed allowing quantification of both effects in different tissues, notably pure water in a gel structure, and porcine muscle and fat tissue. The temperature varied from 28 to 44 degrees C, a range significant for hyperthermia applications. Good agreement with results from the literature was obtained for water. Porcine muscle tissue behaves like water. Its screening constant varies linearly at a rate of 0.97 10(-8) (degree C)-1 and the effects of temperature-induced changes of the susceptibility constant are negligible for muscle thermometry applications. The PRF-temperature relation in fat tissue, however, is almost completely determined by susceptibility effects.

This paper shows the capability of the finite difference time domain (FDTD) method to predict the interaction between the human body and the electromagnetic field generated by a cordless telephone. Both the influence of the human head on the performance of the cordless telephone and the energy deposited in the human head have been determined. The interaction has been evaluated for a simple dipole model and for an accurate “box” model of the telephone. The FDTD method is a versatile method for refining the antenna model. The results show that the use of the accurate box model in combination with a realistic model of the head derived from a nuclear magnetic resonance image is a prerequisite for accurate determination of the near fields induced in the head. The total amount of power absorbed in the head has been compared to the radiated power. From our calculations we found that about 15% of the antenna input power at 900 MHz is absorbed in the head.

In the standard derivation of Jaccard's theory of the electric properties of ice, no fundamental distinction is made between bound and free charges. This leads to some didactical problems like the ad hoc introduction of the so-called 'configuration vector' Ω. However, when the two types of charges are distinguished, it becomes clear that Ω is redundant and proportional to the polarisation density. We also show that Jaccard's formulation contains a wrong formula for the electric susceptibility and that the correct Φ factor can be derived from a straightforward kinetic approach (which Jaccard failed to do).