Electronic devices are increasingly being implanted inside the human body to improve medical diagnostic or therapeutic efficacy. Fine tuning these devices requires communications with medical professionals, yet direct communications are problematic due to the locations of these systems within the body. A growing trend has been to communicate with these systems by means of wireless RF transceivers. Care must be taken when designing an antenna for such systems as canonical antenna design methods no longer apply. Compensation for detuning effects must also be considered due to the nature of the biocompatible materials used and the proximity of constituent body materials. A framework for addressing these issues and a method for retuning implanted antennas is presented here.
Designing antennas for implantable systems presents many problems. Implantable systems are small in volume and often with such unique shapes that standard antenna design techniques do not often apply. Antennas typically are forced to take on the conformal shape of the implant case, and designing antennas based on these smaller dimensions becomes difficult as they begin to rival a fraction of a wavelength of the Medical Implant Communications Service (MICS) frequencies (402 to 405 MHz). Antennas at these smaller dimensions have difficulty achieving the intended gain, bandwidth, and directivity parameters for an application. The effects due to these mechanical constraints are unavoidable and must be ameliorated as best as possible.
Other antenna performance issues in implantables, however, are addressable. Canonical methods used to design antennas usually address free space operation with no surrounding materials to cause complications. All the near-field electrical and magnetic fields remain undisturbed and thereby yield far-field patterns as one would expect. But antennas mounted on systems implanted in the body come in direct contact with constituent body materials that can provide deleterious effects to the expected antenna performance (Fig. 1)
Figure 1 illustrates that body dimensions can have a significant affect in achieving a desired communication distance. A 2-m communication distance is typical for implantable systems, and meeting this criterion for a person with a larger body mass will be more difficult. Reception at the opposite side of the body is also in question as more of the body needs to be traversed.
Overall, the main concern with these devices is maximizing RF coverage, and attempts to boost coverage by increasing output power of the transceiver can shorten battery life and potentially lead to an earlier than expected explant, which is undesirable. Increases in RF power also cannot be made without limit to meet desired communication distances. Regulatory bodies, such as the United States Federal Communications Commission (FCC) and the European Telecommunications Standards Institute (ETSI), place limits on the radiated power from an implanted system at 25 W effective radiated power (ERP). A design methodology can be employed that addresses these antenna and communication shortcomings, combined with designing for smaller dimensions and conformal shape limitations, as will be described below.
Losses in performance come about from interactions with the properties of the body materials. Body materials have properties that act like a combination of a conductor and a dielectric. The losses from these properties can be considered in two different contexts: reflections and direct losses.
Reflections of propagating signals take place at the interfaces between materials. Since the dielectric constants of the body materials change abruptly at these interfaces (muscle/fat, fat/skin, etc.), the majority of the RF wave will be reflected internally with only the remainder transmitted through the material and outside the body.
These reflected waves then propagate through the internal portion of the body where they suffer a fair amount of loss. These losses can be characterized as a sum of losses due to conductivity of the material and the losses due to its dielectric properties. Conductive losses are due to the finite conductivity of the body materials themselves. Dielectric losses come about through the movements of the polarized molecules within the material (e.g., water, sodium chloride, etc.) that oscillate with the electric field variations. These molecules are mutually bound by intermolecular forces that oppose a resistance to the oscillations where energy is lost.
Special antenna design techniques are required to protect the electrical field as much as possible from these dielectric effects. Antennas can be designed that enclose the electric field mostly internal to the antenna itself such that exposure to body materials causes little effect. Antennas can also be designed such that a magnetic element is the dominating radiator (magnetic dipole)magnetic fields are not disturbed in the presence of dielectric materials. These techniques also need to be employed without significantly increasing the size of the antenna.
Despite design attempts to completely protect the electric field, invariably there will be some interaction with the surrounding dielectric material. One of the undesirable effects of the presence of this dielectric material is the detuning of the antenna resonance. If the dielectric material has high permittivity, it will slow the propagation of waves in its vicinity. This decrease in speed is seen as a shorter wavelength, thereby decreasing the overall antenna resonant frequency (Fig. 2). This effect, despite design attempts to protect the electric field, can also cause subtle shifts in the antenna bandwidth and the impedance at the center frequency.
Impacting this problem further is the ever-changing nature of the materials within the body. Changes in body and layer dimensions, combined with even slight shifts in implantable system location, can vary the above mentioned frequency effects over time, requiring the need of a frequency-correcting system.
Exacerbating this detuning effect even further is the nature of the dielectric materials used in the antenna design itself. Since the antenna is implantable and exposed to the human body, the dielectric must be fabricated from a biocompatible material. While biocompatible dielectrics do exist, the manufacturing process for them cause a variance in the permittivity value on the order of +/- 10 percent. This variance can lead to an initial tuning offset before the body effects are taken into account.
The key to retuning and matching an antenna is determining its return loss in the frequency band of interest. One needs to be able to evaluate the changes in frequency performance as the circuits are adjusted. Figure 3 shows a generalized diagram of an antenna/transceiver system with tuning and matching capabilities.
Continue to page 2
Most MICS-band transceivers have separate ports for the transmitter and receiver, which require an external diplexer circuit to use a single antenna. Active tuning and matching of the antenna can now be achieved by a loop feedback system. Inputs into the tuning/matching circuitry are derived from the return loss measurement in the transmitter path. Since the diplexer and passive match circuits in each of the transmitter and receiver paths provide a balanced configuration at the diplexer common port, the easiest way to tune and match the antenna to the common port is by utilizing the signal from the transmitter itself. Once the system is tuned and matched properly to the transmitter, the receiver will maintain its match due to the diplexer and passive match designs.
To achieve the necessary retuning effect, specific circuitry, usually of the variable capacitance type, can be employed in conjunction with the transceiver circuitry driving the antenna to bring the center frequency and impedance match at the center frequency back into desired ranges. A common way to enact variable capacitance is through the use of a varactor diode. Different effective capacitances can be achieved through the voltage used to reverse bias the varactor diode junction. Since the extents of the capacitance from this device would be known, the passive components making up the tuning and matching networks could be selected to achieve all the required frequency shift and impedance modifications necessary to account for variances caused by the human body effects and material variances listed above.
A more specific transceiver system appears in Fig. 4. An optimal tuning/matching solution with this system would involve using three separate variable capacitances, C1, C2, and C3. The matching and tuning networks would each have their own respective variable capacitance, with an additional capacitance at the transmitter output. This topology is similar to what is implemented by existing MICS transceiver device companies, such as Zarlink Semiconductor, Inc.
A typical tuning and matching scenario would proceed as follows: as the transceiver transmitter outputs a desired operational signal in the MICS frequency band, the antenna tuning capacitance (C3) would be varied across its entire range until a maximum voltage is found at the transmitter output, as evaluated through the MUX and ADC for its peak value. This maximal signal would indicate when the antenna is tuned properly to that frequency of operation. Once the tuning has been accomplished, then a similar algorithm would be employed to sweep the variable capacitance in the matching network (C1) across its entire range while observing the maximum voltage seen at the transmitter output when transmitting to improve impedance match. Capacitance C2 can also be adjusted to further improve the transmitter match.
Designing antennas to minimize RF attenuation and detuning effects, and constructing electronic circuitry and software to retune these antennas to the desired operating frequency and impedance match is imperative. The design of the antenna and the design of the electronic circuitry/software should not be thought of as separate design efforts. These components are intimately related and should be developed as a comprehensive subsystem within the overall implant electronics.