Transcutaneous Energy Transfer System for Powering Implantable Biomedical Devices

Transcription

1 Transcutaneous Energy Transfer System for Powering Implantable Biomedical Devices T. Dissanayake 1, D. Budgett 1, 2, A.P. Hu 3, S. Malpas 2,4 and L. Bennet 4 1 Auckland Bioengineering Institute, University of Auckland, Auckland, New Zealand 2 Telemetry Research, Auckland, New Zealand 3 Department of Electrical and Computer Engineering, University of Auckland, Auckland, New Zealand 4 Department of Physiology, University of Auckland, Auckland, New Zealand Abstract Time varying magnetic fields can be used to transfer power across the skin to drive implantable biomedical devices without the use of percutaneous wires. However the coupling between the external and internal coils will vary according to orientation and posture. Other potential sources of power delivery variations arise from changes in circuit parameters and loading conditions. To maintain correct device function, the power delivered must be regulated to deal with these variations. This paper presents a TET system with a closed loop frequency based power regulation method to deliver the right amount of power to the load under variable coupling conditions. The system is capable of regulating power for axially aligned separations of up to 10mm and lateral displacements of up to 20mm when delivering 10W of power. The TET system was implanted in a sheep and the temperature of implanted components is less than 38.4 degrees over a 24 hour period. Keywords Magnetic field, coupling, Transcutaneous Energy Transfer (TET) I. INTRODUCTION High power implantable biomedical devices such as cardiac assist devices and artificial heart pumps require electrical energy for operation. Presently this energy is provided by percutaneous leads from the implant to an external power supply [1]. This method of power delivery has the potential risk of infection associated with wires piercing through the skin. Transcutaneous Energy Transfer (TET) enables power transfer across the skin without direct electrical connectivity. This is implemented through a transcutaneous transformer where the primary and the secondary coils of the transformer are separated by the patient s skin providing two electrically isolated systems. A TET system is illustrated in figure 1. The electromagnetic field produced by the primary coil penetrates the skin and produces an induced voltage in the secondary coil which is then rectified to power the biomedical device. Compared to percutaneous wires, TET systems become more complex to operate under variable coupling conditions as it result in a variation in power transfer [2]. One source of variation in coupling is due to posture changes of the patient causing variation in the alignment between the primary and the secondary coils. The typical separations between the internal and external coils are in the range of 10-20mm. If insufficient power is delivered to the load then the implanted device will not operate properly. If excessive power is delivered, then it must be dissipated as heat with the potential for causing tissue damage. Therefore it is important to deliver the right amount of power matching the load demand. DC Supply Primary Coil Power Converter Controller Skin Magnetic coupling Fig. 1 Block diagram of a TET system Secondary Coil Pickup Load Power feedback Power can be regulated either in the external or the implanted system. However, regulation in the implanted system results in dissipation of heat in the implanted circuitry [3]. Furthermore, it also increases the size and weight of the implanted circuitry therefore power regulation in the external system is preferred over the implanted system. There are two main methods of regulating power in TET systems, magnitude and frequency control methods. In the case of magnitude control, input voltage to the primary power converter is varied in order to vary the power delivered to the load. This method of control is very common in TET systems however it does not take into account the miss-match of the resonant frequency of the secondary resonant tank and the operating frequency of the external power converter. This miss-match in frequency reduces the power transferred to the load, consequently, a larger input voltage is required which results in a reduction in the overall power efficiency of the system. Frequency control involves varying the operating frequency of the primary power converter to vary the Chwee Teck Lim, James C.H. Goh (Eds.): ICBME 2008, Proceedings 23, pp ,

2 236 T. Dissanayake, D. Budgett, A.P. Hu, S. Malpas and L. Bennet power delivered to the load. Depending on the actual power requirement of the pickup load, the operating frequency of the primary power converter is varied so the secondary prick-up is either tuned/detuned, thus the effective power delivered to the implantable load is regulated [4]. The system discussed in this paper uses frequency control to control power delivery to the load, and a Radio Frequency (RF) link is used to provide wireless feedback from the implanted circuit to the external frequency controller. II. SYSTEM ARCITECHTURE The TET system is designed to deliver power in the range of 5W to 25W. Figure 2 illustrates the architecture of the overall system. A DC voltage is supplied to the system with an external battery pack. A current fed push pull resonant converter is used to generate a high frequency sinusoidal current across the primary coil. The magnetic coupling between the primary and the secondary systems produces a sinusoidal voltage in the secondary coil which is rectified by the power conditioning circuit in the pickup to provide a stable DC output to the implanted load. As shown in figure 2, a DC inductor is added to the secondary pick up following the rectifier bridge in order to maximize the power transfer to the load. The DC inductor aids to sustain a continuous current flow in the pick up [5]. Push-pull resonant Frequency controller V ref Digital analogue converter External transceiver Primary resonant tank C p L p Skin Secondary resonant tank L s C s Magnetic coupling RF Communication channel Fig. 2 System architecture Biomedical load V dc Internal transceiver Two nrf24e1 Nordic transceivers are used for data communication. The DC output voltage of the pickup is detected and transmitted to the external transceiver. The external transceiver processes the data and adjusts the duty cycle of the output PWM signal in order to vary the reference voltage (V ref ) of the frequency control circuitry. The PWM signal is passed through a Digital to Analogue Converter (DAC), in order to obtain a variable reference voltage. This variable reference voltage is then used to vary the frequency of the primary resonant converter which in turn varies the power delivered to the implantable system. The response time of the system is approximately 360ms. A. Frequency controller The frequency controller employs a switched capacitor control method described in [7]. The controller varies the overall resonant frequency of the primary resonant tank in order to tune/detune to the secondary resonant frequency. The frequency of the primary circuit is adjusted by varying the effective capacitance of the primary resonant tank. This is illustrated in figure 3. V IN C V1 S V1 L 1 L 2 S 1 S 2 Secondary resonant Primary resonant tank tank C V2 S V2 Fig. 3 System based of primary frequency control [4] Inductor L P, capacitor C P and switching capacitors C V1 and C V2 form the resonant tank. The main switches S1 and S2 are switched on and off alternatively for half of each resonant period and changing the duty cycle of the detuning switch S V1 and S V2 varies the effective capacitances of C V1 and C V2 by changing the average charging or discharging period. This in turn will vary the operating frequency of the primary converter. Each C V1 and C V2 is involved in the resonance for half of each resonant period. The variation in reference voltage (V ref ) obtained from the DAC is used to vary the switching period of these capacitors. This method of frequency control maintains the zero voltage switching condition of the converter while managing the operating frequency. This helps to minimize the high frequency harmonics and power losses in the system. As shown in figure 3 the pickup circuitry is tuned to a fixed frequency using the constant parameters L S and C S. The operating frequency of the overall system is dependent on the primary resonant tank which can be varied by changing the equivalent resonant capacitance [6], therefore the tuning condition of the power pickup can be controlled. C P L P L S C S Load

3 Transcutaneous Energy Transfer System for Powering Implantable Biomedical Devices 237 III. EXPERIMENTAL METHOD A prototype TET system was built and tested in a sheep. The internal coil and the resonant capacitor were Parylene coated and encapsulated with medical grade silicon to provide a biocompatible cover. The total weight of the implanted equipment was less than 100g. As illustrated by the cross sectional view in figure 4, thermistors were attached to the primary and the secondary coils to measure the temperature rise caused by the system in the surrounding tissue. Thermistor 1: Placed on top of primary Thermistor 2: Placed under the skin Thermistor 3: Placed on the muscle side Thermistor 4: 1cm from the secondary coil Thermistor 5: 2cm from the secondary coil Thermistor 6: Near the subcutaneous tissue near the exit of the wound. Prior to experimentation, the thermistors were calibrated against a high precision FLUKE 16 Multimeter temperature sensor and a precision infrared thermometer. 6 Primary coil Secondary coil Healthy tissue Subcutaneous tissue 1cm 4 5 2cm also put on the site of the wound to reduce infection. Following the surgery the sheep transferred to a crate where it was kept over a three week period. The primary coil was placed directly above the secondary coil and held on the sheep using three loosely tied strings. A PowerLab ML820 data acquisition unit and Labchart software (ADInstruments, Sydney Australia) was used for continuous monitoring of the temperature, the output power to the load and the variation in input current of the system during power regulation. The data acquisition was carried out at a frequency of 10 samples per second. IV. EXPERIMENTAL RESULTS Experimental results were obtained for delivering 10W of power to the load when the system was implanted in sheep. Figure 5 illustrates the closed loop controlled power delivered to the load over a period of 24 hours. The input voltage to the system was 23.5V. The controller is able to control the power to the load for axially aligned separations and lateral displacements between 10mm to 20mm. Beyond this range the coupling is too low for the controller to provide sufficient compensation, and delivered power will drop below the 10W set point. Evidence of inadequate coupling can be seen at intervals in Figure 5.Variation in input current reflects the controller working to compensate for changes in coupling using frequency variation between 163 khz (fully detuned) and 173 khz (fully tuned). When the coupling between the coils is good, the primary resonant tank is fully detuned in order to reduce the power transferred to the secondary. When the coils are experiencing poor coupling, the primary resonant tank is fully tuned to increase the power transfer between the coils. Fig. 4 The placement of the temperature sensors Prior to the surgery all implantable components were sterilized using methanol gas. The sheep was put under isoflurane anesthesia and the right dorsal chest of the sheep was shaved. Iodine and disinfectant was applied over the skin to sterilize the area of surgery. Using aseptic techniques a 5 cm incision was made through the skin on the dorsal chest. A tunnel was created under the skin approximately 20 cm long and a terminal pocket created. The secondary coil and the thermistors were placed within this pocket. The thickness of the skin at this site was approximately 10mm. The secondary coil was then sutured in place and the power lead from the coil and leads of the thermistors were tunneled back to the incision site and exteriorised through the wound. The wound was stitched and Marcain was injected to the area of the wound. Iconic powder was Power (W) Graph of closed loop control power at 10 W Output Voltage input current Time (mins ) Fig. 5 Regulated power to the load and the input current to the system Figure 6 shows the temperature recorded from the six thermistors. It takes approximately 20 minutes for the tem Input current (A)

4 238 T. Dissanayake, D. Budgett, A.P. Hu, S. Malpas and L. Bennet perature to reach a steady state after turn-on. The maximum temperature was observed in the thermistor placed under the secondary coil on the muscle side. The maximum temperature observed in this thermistor over the 24 hour period was C. The maximum temperature rise observed was C in the thermistor placed under the skin. The large variation in the primary coil temperature is due to the changes in current through the coil from the frequency control mechanism. When the system is in the fully tuned condition, the current in the primary coil is at a maximum to compensate for the poor coupling. The temperature rise in the thermistors 1cm and 2cm from the secondary coil is well below 2 0 C. Temperature (Celcius) Temperature against time when delivering 10W Time (mins) under skin muscle side 2 cm fro m s ec near wound exit primary surface 1 c m fro m s ec Fig. 6 Temperature profile of the thermistors V. DISCUSSION Although the system performs well at delivering 10W over a 24 hour period, there are short intervals when this power level was not delivered. These intervals correspond to times when the coupling is too low for the controller to compensate. A variety of approaches can be taken to solve this problem. The first is to tighten the coupling limitations to prevent coupling deteriorating to beyond the equivalent limit of 20mm of axial separation. The second approach is to allow occasional power drops on the basis that an internal battery could cover these intervals, (patient alarms would activate if the problem persists). The third approach is to increase the controller tolerance to low coupling. The ability to tolerate misalignments of the frequency controlled system is mainly determined by the systems quality factor (Q value), which is defined by: Q R L (1) Where R is the load resistance, (2f) is the system angular operating frequency, and L is the secondary coil inductance. A larger Q will enable the system to be more tolerant. This benefit is traded off against the need for a more sensitive and faster feedback response from the control system. VI. CONCLUSIONS We have successfully implemented a system that is capable of continuously delivering power to a load in a sheep. The results have been presented for delivering 10W of power to the load with closed loop frequency control technique for a period of 24 hours. The external coil was loosely secured to lie over the region of the internal coil and subjected to alignment variations from a non-compliant subject. The maximum temperature observed in this system is C on the thermistor placed on the muscle side under the primary coil. The maximum temperature rise was C on the thermistor placed under the skin. REFERENCES 1. Carmelo A. Milano, A.J.L., Laura J. Blue, Peter K. Smith, Adrian F. Hernadez, Paul B. Rosenberg, and Joseph G. Rogers, Implantable Left Ventricular Assist Devices: New Hope for Patients with End stage Heart Faliure. North Carolina medical journal, (2): p C. C. Tsai, B.S.C.a.C.M.T. Design of Wireless Transcutaneous Energy Transmission System for Totally Artificial Hearts. in IEEE APPCAS Tianjin, China. 3. Guoxing Wang, W.L., Rizwan Bashirullah, Mohanasankar Sivaprakasam, Gurhan A. Kendir, Ying Ji, Mark S. Humayun and James D.Weiland. A closed loop transcutaneous power transfer system for implantable devices with enhanced stability. in IEEE circuits and systems Ping Si, P.A.H., J. W. Hsu, M. Chiang, Y. Wang, Simon Malpas, David Budgett Wireless power supply for implantable biomedical device based on primary input voltage reglation. 2nd IEEE conference on Industrial Electronics and Applications, Ping Si, A.P.H., Designing the DC inductance for ICPT Power pickups Ping Si, A.P.H., Simon Malpas, David Budgett, A frequency control method for regulating wireless power to implantable devices. IEEE ICIEA conference, Harbin, China, Author: Thushari Dissanayake Institute: Auckland Bioengineering Institute Street: 70, Symonds Street t.dissanayake@auckland.ac.nz

5 Transcutaneous Energy Transfer System for Powering Implantable Biomedical Devices 239 Author: David Bugett Institute: Auckland Bioengineering Institute Street: 70, Symonds Street Author: Patrick Hu Institute: University of Auckland Street: 38, Princess Street