Figure 5.33: The dependency of the resonance frequency on the surrounding tissues.
side of 2.15 cm and a volume of 10 cm3. This is four times the thickness of our model implant. In order to investigate the very local peaks of the absorption, SAR was calculated over three different volumes, 10 g, 1 g and 1 mg, where the 1 mg corresponds to a cube with a side of 1 mm. The results are presented in Table 5.5 and the table is normalized to 1 W input power into the antenna. If 1 mW, or 0 dBm, is used as the power to the antenna, all values should be divided by a factor of 1000. All the antennas are then below the Swedish limit. Devices with a mean output power below 0.02 W do not need to be tested in Sweden, as they are not assumed to give a SAR above the limit [63]. Our simulations verify this assumption in the tested antennas, as the antenna with the highest SAR, the un-insulated wire antenna, gives a SAR of 1.0 W/kg in 10 g when fed with 20 mW of continuous power. The FCC has a proposal [64] that devices worn at the body with an output below 2 mW should be exempt from licensing. This would give the un-insulated wire antenna a SAR of 0.6 W/kg over 1 g when fed with 2 mW of continuous power. This is below the FCC limit of 1.6 W/kg.
5.7 Conclusion
The result from this chapter is that the circumference type antennas and the magnetic coil antenna are the most interesting ones for medical implant appli-cations. The patch antenna is not very well suited for placement inside a lossy
68 CHAPTER 5. ANTENNA DESIGN SAR (W/kg)
Antenna 10 g 1 g 1 mg
Un-insulated Wire 52 300 7900
Isolated Wire 4.3 12 47
Circumference Wire 8.1 14 150 Circumference Plate 8.3 22 260 Circumference PIFA 8.0 21 240 Table 5.5: SAR levels of the implanted antennas.
matter, unless it is well insulated from the matter. It will then loose the benefit of having a small volume.
The circumference antenna is more conforming to an existing antenna shape than the coil antenna, which, on the other hand, has a smaller total volume.
The theoretical calculations point in the direction that the magnetic antenna should have a larger gain in the case of a perfectly conducting antenna coil.
On the other hand, it has larger resistive losses in the coil wire in an actual implementation than the wire type antennas.
Chapter 6
Influence of Patient
When we place a medical implant with an antenna inside a patient, the antenna will be affected by the immediate surroundings. Thus, the antenna behaves differently if placed in an arm, deep in the abdomen or just beneath the skin in the chest. In addition to this there will be a dependency on the surrounding tissue type, for example variations in the subcutaneous fat layer. This layer varies in thickness between patients and varies also over time, when a patient gains or loses weight. As will be shown, the far-field from the antenna is affected by the patient’s size, body shape and position. We have investigated some of these effects by numerical simulations for an implant placed in the chest region of a human. This has been done in order to characterize the magnitude of the variations and to be able to set levels for the excess loss in the link budget calculations. We have also investigated the influence of the shape of the body depending on age and sex.
Movements of the patient change the immediate surrounding of the im-planted antenna. The MICS frequency band, 402-405 MHz, corresponds to a wavelength of approximately 74 cm in air and approximately 9 cm inside the body. The body surface is in the near-field of the implanted antenna. Any change of the permittivity or conductivity of materials placed in the near-field of an antenna changes its radiation characteristics. Thus, a change in posture changes the far-field pattern and affects the radio channel between the medical implant and the external base station. This variation of the channel corresponds to a kind of slow fading. It was known at an early stage of the project that the posture of the patient influences the far-field of an implanted antenna [65].
The gain, directivity and efficiency of an antenna were defined in Chapter 5. The antenna has an efficiency factor η, 0 ≤ η ≤ 1,which is a measure of how much power is lost in ohmic losses in the antenna and in the body. If the antenna is lossless then η = 1, which is impossible to get for an implanted antenna. The antenna is taken to be perfectly matched. Thus, reflection losses due to mismatch are not included in the gain. In this chapter the gain is determined in the far-zone of the body, which is in air, and this makes the gain and the directivity independent of the origin. The gain taken from simulations
69
70 CHAPTER 6. INFLUENCE OF PATIENT in this thesis is scaled in dBi, or decibels relative an isotropic antenna. An isotropic antenna is a theoretical construction that radiates equally well in all directions.
6.1 Method
We built the different phantoms as described in Appendix E. As excitation we used the wire antenna connected to our standard pacemaker model, or the circumference wire antenna. The simulations were done in SEMCAD, as in the previous chapter. As absorbing boundary condition Mur [66] was used in order to reduce the simulation time.