An Alternative Approach of Operating a Passive RFID Device Embedded on Metallic Implants

Transcription

1 An Alternative Approach of Operating a Passive RFID Device Embee on Metallic Implants Xiaoyu Liu, Ravi Yalamanchili, Ajay Ogirala an Marlin Mickle RFID Center of Excellence, Department of Electrical an Computer Engineering University of Pittsburgh Pittsburgh, USA xil56@pitt.eu (Xiaoyu Liu) Abstract This paper presents a metho of using volume conuction for communication with implante passive RFID evices. Generating mainly electric fiel through human tissue, this metho can reuce the interference with metallic implants when RFID evices are attache to the implants. Exterior an interior electroes are esigne for efficient energy an signal transmission. Experiments with a saline box are conucte to stuy the signal attenuation through biological tissue an the feasibility of using volume conuction for RFID evices close to in vivo metallic materials. Keywors-implante RFID; interference I. INTRODUCTION volume conuction; metallic Raio frequency ientification (RFID) technology is a wireless communication technology that uses an electromagnetic fiel at raio frequency for information exchange. Typical RFID systems inclue interrogators (reaers) an tags an the link between them is wireless. The major application of RFID technology is for ientification an tracking purposes such as Wal-mart type tags [1]. Due to its small size an low cost, one potential application of RFID evice is embeing sensors or boy implants such as artificial joints (Fig. 1). Such application can provie a convenient way for octors an surgeons to reuce counterfeiting an obtain information from patients who have experience joint replacement surgeries, with little concern about the loss or error in searching information. Furthermore, if integrate with sensors, such implante RFID evices can also be use for monitoring the status of the implante joints, an thus help to iagnose. The eman for such implante RFID evices has grown rapily in recent years. Such eman will increase as the population ages. By the year 2030, it is estimate that the number of knee replacements will increase from 450,400 to 3.48 million [2]. One major concern is the tag performance with current RFID systems, since the artificial implante joints are mae of metal. There has been consierable research iscussing the interference between RFID systems an metallic materials [3] [4]. How interference affects the RFID systems is a case by case problem epens on position, size an structure of the material as well as the material itself. For stuying the possibility to operate an implante RFID evice using traitional tags in such a conition, experiments on reaing a joint embee RFID tag are reporte in this paper. The experiment is performe using a stanar Magellan MARS-2 reaer with the setup as shown in Fig. 2. We observe that no reaing of the tag is seen to be recore. Even when move far away from the joint, the embee tag is only able to be rea with low efficiency meaning that uring n number of times reaing only a small fraction is successful. As biological tissue attenuates HF signals more than the air, we believe that the traitional wireless methos to operate a RFID tag embee on metallic materials are problematic. An alternative approach is require to operate the RFID evices in such a conition. Stuies have shown that ionic fluis within biological tissues are capable of conucting electrical current which, when intentionally manipulate, can be use to transmit information an energy. This metho of using the boy as a conuctor, known as volume conuction, has been use to transmit information from a sensor implante within the leg of a caaver to perform mechanical measurements [5], an sen information using a boy bus escribe in a Microsoft patent [6]. It has also been use experimentally to recharge the battery of an implantable evice to eliminate battery replacement surgeries [7]. Since volume conuction uses mainly electric fiel in conucting power, it has fewer chances to inuce big ey current in metals, an therefore is more power efficient an has less interference in metallic environment than using the wireless way. Using volume conuction also eliminates the nee of antenna, which in turn makes it more flexible in the size an shape of implante RFID evice. Base on these research results an implementations, we propose a metho of using volume conuction to power an communicate with a passive RFID evice that is embee with in vivo metallic materials. In this paper, we focus on creation of an RFID volume conuction system an iscussion about its feasibility in use with metallic environments. Three sections are on the esign of the system incluing the volume conuction system overview, electroe esign an RFID evice matching. Finally, in section V, the experiment of successfully reaing a passive RFID evice that is embee on a real artificial joint has shown the feasibility of using our propose metho in reucing interference with metallic environments.

2 B. The saline box The saline box is built for stuying signal attenuation through biological tissue where volume conuction is implemente. The box is shown in Fig. 4. The box is mae from 0.09 cm thick acrylic boars. The imension of the box is mae to be 30 cm x 30 cm x 28 cm in orer that sufficiently large transmission ranges can be covere. The exterior boy electroes are attache on the inner sie of one of the walls making them contact the surface of the saline. The interior electroes are attache to a movable panel that prouces the istance to the exterior electroes. Stainless screws are use for fastening the electroes to the box an the panel, an also serving as external terminals. Figure 1. Artificial joint an the possible locations for RFID implants. Figure 2. Reaing an embee tag place close to joint implants wirelessly. II. SYSTEM OVERVIEW A. Volume Conuction System The propose volume conuction system in this paper relies on the same principle as in previous references [6-7]. The system is illustrate in Fig. 3. The source is connecte to the boy through two external electroes attache to the surface of the skin. An electric fiel is generate when a voltage is prouce across the external electroes. Two internal electroes connecting the RFID evice are implante to interact with the electric fiel to capture the signal. The signal capture by the internal electroes is use to power the RFID chip as well as for communication. The RFID evice backscatters information through the inverse path but with the same principle. Figure 4. The saline box. C. Electroes Electroe esign is one of the most important consierations for volume conuction systems. Various electroes in shape, size an patterns are esigne to fulfill ifferent applications [8] [9]. In this paper, for simplicity, ipole electroes with a roun shape are use. The electroes are mae of stainless steel to prevent corruption in saline. The sizes of the electroes range from 4cm to 1cm in iameter, with a through hole for screws at each center (Fig. 5). The thickness of the electroes is 1mm, making it suitable for implantation. Figure 5. Electroes in ifferent sizes. Figure 3. Volume conuction system. III. SIGNAL ATTENUATION One major concern using volume conuction is that biological tissue attenuates signals, especially at higher frequencies [10]. The more signal it attenuates, the less power

3 that an implante RFID evice can receive. Therefore, it is crucial to know how much signal is attenuate when using ifferent sizes of electroes with ifferent istances in between electroes. Such experiments are conucte using the saline box in Fig.4. Due to the limitation of the surface area of skin, the exterior electroes are selecte to be 4cm in iameter with 5cm between each center. The exterior electroes are place an fixe as a pair at the center of the wall at one sie of the saline box (as shown in Fig. 3). This makes them equal in istance to the sie walls of the box an the bottom. The interior electroes are aligne with the exterior electroes in such a manner that their flat surfaces are in parallel an they have the same height to the bottom of the saline box from their centers. Since the in vivo space is limite for embeing implantable RFID evices, the sizes of interior electroes are then selecte to be in the range between 2.5 cm to 1 cm in iameter with 0.5cm ecrement. The istance between the interior electroes has a maximum of 5cm. The electrical signal source is a Rohe & Schwarz SMJ 100A function generator which connects to the exterior electroes through an RF cable. The function generator outputs an 18Bm sine wave at 13.56MHz, the operating frequency of the RFID reaer. The RMS voltage across the exterior electroes an the RMS voltage across the interior electroes are measure using a Tektronix DPO 7254 igital phosphor oscilloscope. The experiments use 0.18% saline, which is the type of saline that follows the requirement specifie in [11]. The measurement results are shown in Fig. 6 an Fig. 7 where the signal attenuation is efine as in equation (1). The letters L an D in the figures represent the istance between the center of the interior electroes an the iameter of the interior electroes respectively. The saline epth is the thickness of saline that RF signal goes through in the experiments. V A = 20 log 10 V input output (1) Figure 7. Signal attenuation through saline using ifferent sizes (D) of interior electroes with 3cm istance between each center. As can be seen from the above figures, the larger interior electroes with larger istances between them ten to reuce the signal attenuation as the epth of saline increases. However, as the epth increases, say to 5cm, significant attenuation occurs. This is because the electric fiel is extremely weak at greater epths making larger electroe sizes an istances not able to capture sufficient flux. At such epths, not enough power can be transmitte to the implante RFID evices. IV. MATCHING CIRCUIT Impeance matching is important in electrical circuits to obtain maximum power transfer between a source an a loa. In orer for maximum power transfer to occur, the impeance of the source an loa are require to be complex conjugates of each other, i.e., R +jx an R jx, where R is the real part an X is the imaginary part of the impeance. Matching networks are also require in RF systems for maximizing power an signal transmission efficiency. In this volume conuction system, matching circuits are use at both the external an internal sies for optimum performance. The iagrams of the matching circuits are shown in Fig. 8. Figure 8. The matching circuit for volume conuction system. Figure 6. Signal attenuation through saline using 1 cm interior electroes with ifferent istances between electroe center (L). A. Matching the saline to the reaer Matching the impeance of the saline to the reaer is require to reuce the signal reflecte back to the reaer ue to

4 any impeance mismatches, thus maximizing the energy transmitte to the saline acting as a loa on the generator. The impeance of the saline is the impeance seen from the exterior electroes into the saline, which is usually greater than the impeance of the reaer. An LC circuit [12] is ae between the reaer an the saline as inicate in Fig. 8. Solving the equation that the equivalent impeance seen from the external matching network is equal to the impeance of reaer, the values of the inuctor an capacitor that can match the saline to the reaer can be calculate from following equations: C 1 1 = ωr 0 R 0 R R L ( R R ) RL 0 L X L L1 = (3) ω where ω is the angular frequency, R0 is the impeance of reaer (50 ohm), an RL an X L are the real an imaginary part of the saline impeance. The saline is measure to be 21.6 j0.61 ohm by an Agilent 8753S network analyzer, which in turn gives a 270 pf (pico Fara) capacitance an 280 nh (nano Henry) inuctance for the matching network. B. Matching the RFID evice Matching the impeance of RFID evice to the saline is much more complicate because the impeance from the interior electroes to the reaer is very sensitive to the installation of the interior electroes. The saline presents a 3 imensional loa which is thus affecte by the location of the electroes an the imensions of the box holing the saline. To some extent, the acrylic box itself is a part of the bounary conitions which can be ignore in this case because the matching is being performe empirically. It is also ifficult to measure or calculate the impeance of saline from the internal sie when reaer is connecte. Therefore, it is impractical to match the impeance of the RFID evice to a specific value that can be calculate a priori. The way we propose to etermine the matching impeance is to maximize the voltage across the electroes of the RFID evice. A larger loa voltage means more power as long as there is enough voltage to turn on the RFID evice. It is important to note that while power is to be optimize, there must be sufficient voltage to turn on the transistors in the RFID chip. Thus, the actual matching circuit is assume in this case to be the one that prouces the largest voltage across the electroes where the power is assume to be sufficient to rive the silicon chip of the RFID tag. Such a conition can be easily satisfie ue to the moern manufacturing technology which reuces the power level of the RFID chips from mw to µw. From the internal to external, the part of circuit from saline to the reaer can be seen as a source with impeance. The impeance is extremely large because the current at internal L (2) sie to be very low. Thus, maximization of the loa voltage is the metho to match the loa impeance to the source impeance. Because the impeance of the RFID evice, in terms of Z, is capacitive, inuctive impeance R 2 + jx 2 is ae in parallel with Z as inicate above to prouce a conjugate impeance. The problem then becomes one of optimizing the impeance at the loa which is given by the following equation: ( R + jx )( R + jx ) R2, X 2 > max (4) R + jx + R + jx where; R an X are the real an imaginary parts of the RFID chip impeance. Given the Q factor [ref] of the matching impeance, one can calculate the optimal solution for X 2 as 2 2 ( R + X ) Q X 2 = (5) R + X Q The real part of the matching impeance is then calculate by iviing X 2 by Q, an the inuctance of the circuit can be obtaine from iviing X 2 by the angular frequency. The impeance of the RFID evice is 39-j800 ohm by measurement using the network analyzer. For size consieration at the implant, surface mount inuctors are use. Comparing the available parts in the market, we have chosen a Q = 30 for calculation which gives an inuctance of 9.4µH. In the experimental system, consiering capacitance ae by the electroes an also the availability of parts in iscrete units, a 10 µh inuctor is use in this prototype. V. EXPERIMENT To test the feasibility of using volume conuction to operate an RFID evice embee on a metallic implant, experiments in reaing an RFID chip with electroes attache is performe. First, the prototype of an embee RFID evice is built on a printe circuit boar (PCB) to accommoate both the RFID chip an the matching circuit. Due to the limitation of uner-skin volume, the size of the PCB is require to be as small as possible. Therefore the initial size of the electroes is chosen to be 1cm in iameter with 2cm in istance between (Fig. 9a). The prototype PCB is then place to attach to the right sie on the surface of the metal (Fig. 9b). The equipment use in this experiment inclue an HF RFID reaer with operation frequency of 13.56MHz (Magellan MARS-2), the saline box an its matching circuit, an actual artificial joint, the RFID chip prototype boar with matching an an oscilloscope. The RFID chip is esigne to allow 8Kb of memory with ata rate up to 96Kb/s. This is sufficient for storing the information pertaining to the joint an relevant meical information for the patient. A small resistor is connecte in series with the saline at the external sie. By measuring the voltage across the resistor, the current flowing into the saline can be measure. Then the power transferre to

5 the saline can be calculate an any safety evaluation can be performe. Fig. 10 shows the experimental setup. In the experiment, the power of the RFID reaer is set to be the maximum, i.e., the same level as the interference test at the beginning when using a wireless antenna. The artificial joint is tape onto a movable panel for test purposes allowing ifferent saline epths. Figure 11. Attenuation with an without a metallic joint using the prototype embee RFID evice. a Figure 9. a. The prototype RFID implant an. b. place that the RFID is attache to the joint. b It is believe that greater istances in the saline can be achieve by optimizing the electroes on the reaer sie of the energy transfer. In aition, the ispersion (iffusion) of the energy with an actual knee will be consierable less than the large volume of saline use in this initial experimental setup. The current in the saline uring the test is approximately 76mA, well below the safety level specifie by [13]. This also allows for increase performance within the limits of safety. At this level of current an taking into account that the reaing time for the RFID evice is normally less than 1 secon, we believe that the volume conuction system propose will not cause tissue burn an is safe in human boy. VI. CONCLUSION The experiment results show no severe attenuation increase is encountere when the RFID tag is place in the metallic environment of an orthopeic implant using volume conuction. The success in reaing the prototype embee RFID evice proves the feasibility of operating an implante RFID in the metallic environment of an orthopeic implant. Volume conuction is the best approach for an RFID system to reuce metallic interference when consiering the human boy. Figure 10. The experiment setup. One of the problems with RFID with an air meium in the proximity of metal is reuce performance. This situation is being teste next with the presence of the saline as the meium of transmission with volume conuction. The efficiency with an without the artificial joint is therefore measure an shown in Fig. 11. We see that with the metallic joint in the experiment, the attenuation is only a little larger than with no joint. This means that the metallic interference of RFID system using volume conuction is significantly reuce. With 11.6% less in voltage efficiency (the ratio of receive voltage to supplie voltage) than with no joint, the RFID evice embee on the metal joint can be rea through 1.5cm of saline without ifficulty. The best efficiency for reaing occurs within 1cm of range which is applicable for most people. ACKNOWLEDGMENT The authors woul like to thank the Office of Technology Management of University of Pittsburgh an LB Meical for their support of this research. REFERENCES [1] J. Banks, Past, present an future of RFID, IEEE International Conference on Automation an Logistics, pp , August [2] S. Kurtz, Total knee an hip replacement surgery projections show meteoric rise by Orthopaeic proceures set to continue gaining wiesprea acceptance as means to restore quality-of-life, Annual Meeting of the American Acaemy of Orthopaeic Surgery, March [3] P. Foster an R. Burberry, Antenna problems in RFID systems, in Inst. Elect. Eng. Colloquium on RFID Technology, pp , 1999.

6 [4] D. Dobkin. an S. Weigan, "Environmental effects on RFID tag antennas," IEEE MTT-S International Microwave Symposium Digest, pp , [5] D. Linsey, E. McKee, M. Hull, an S. Howell, A new technique for transmission of signals from implantable transucers, IEEE Trans. Biome. Engr., vol. 45, pp , [6] L. Williams et al (Microsoft Corporation). Metho an apparatus for transmitting power an ata using the human boy. U. S. Patent No. 6,754,472, Jun [7] S Hackworth, M. Sun an R. Sclabassi, Skin-electroe circuit moel for use in optimizing energy transfer in volume conuction systems, IEEE EMBC, pp , 2009 [8] B. Wessel, Analytical an Numerical Optimization of an Implantable Volume Conuction. (Master thesis, University of Pittsburgh, 2004). [9] S. Hackworth, Design, Optimization, an Implementation of a Volume Conuction Energy Transfer Platform for Implantable Devices. (Doctoral issertation, University of Pittsburgh, 2010) [10] P. Cook, "Raio frequency attenuation of saline solutions," IEEE. International Symposium on Electromagnetic Compatibility, pp , [11] American National Stanar. ANSI/AAMI PC-69:2007 Active implantable meical evices-electromagnetic compatibility-emc test protocols for implantable cariac pacemakers an implantable carioverter efibrillators, [12] K. Chang. Hanbook of RF/microwave components an engineering. Wiley-Interscience, NJ, [13] IEEE stanar for safety levels with respect to human exposure to raio frequency electromagnetic fiels, 3kHz to 300 GHz. IEEE St C95.1, 1999