Feature

The Ultra-Low-Power Wireless Medical Device Revolution


Find more content on:

 

Originally Published MEM Spring 2005

IMPLANTABLES

 

Remote management of implantables requires an ultra-low-power, high-performance transceiver with an architecture specially designed for these medical devices.

Peter Bradley

Implantable medical devices (IMDs) have a history of outstanding success in the treatment of many diseases, including heart diseases, neurological disorders, and deafness. Today's aging population is driving wide-scale demand for more-advanced healthcare treatments, including wireless implant devices that can deliver ongoing and cost-effective monitoring of a patient's condition.

Ultra-low-power integrated circuits help drive implantable devices such as hearing aids.
(click to enlarge)

New ultra-low-power radio-frequency (RF) technologies are spurring the development of innovative medical tools, from endoscopic camera capsules that are swallowed, to implanted devices that wirelessly transmit patient health data. Communication links between external programming devices (or base stations) and medical implants are critical to the success of IMDs. The communication link enables a clinician to reprogram therapy and obtain useful diagnostic information.

Historically, low-frequency inductive links (introduced in the early 1970s) have been the most prevalent method of communication. They typically operate in the tens to hundreds of kilohertz range, with data rates of 1–30 Kb/sec. These low-power systems, which can accommodate a small coiled antenna in the IMD, have proven to be robust and suitably reliable.

Antenna size and power limitations result in very low magnetic field strength.
(click to enlarge)

However, antenna size and power limitations in implants result in a very low magnetic field strength for an IMD that is communicating with an external programmer. Therefore, inductive links are short range and often require the external programmer to have contact with the skin of the patient directly over the implant.

To overcome these operating-range and low-data-rate limitations, new ultra-low-power RF technologies are being developed that operate at much higher frequencies, such as in the 433 and 915 MHz industrial, scientific, and medical (ISM) bands and the more recently allocated 402–405 MHz medical implant communication service (MICS) band. RF integrated circuit (RFIC) technology can now offer low power, reduced external component count, and higher levels of integration, which will open new markets for medical device manufacturers.

This article discusses the issues related to remote management of IMDs via the MICS band and presents guidelines for addressing these issues. A brief history of the MICS standard is presented, followed by an outline of the requirements driving RF medical technology. To enable use of the MICS band, medical devices require an ultra-low-power, high-performance transceiver. This article examines a transceiver IC designed specifically for this purpose. The design considerations of implantable transceivers are presented, followed by a brief discussion of the architecture and important design features.

The MICS Band

A sensor communicates with an external processor and transceiver, which relays the necessary information to the implanted muscle or nerve stimulator.
(click to enlarge)

From a regulatory viewpoint, the establishment of the MICS band began in the mid-1990s when Medtronic petitioned the FCC to allocate spectrum dedicated to medical-implant use. After gaining wider industry support, the 402–405 MHz MICS band was recommended for allocation by ITU-R Recommendation SA1346 in 1998. FCC established the band in 1999, and similar standards followed in Europe.

The allocation of this band supports the use of longer- range (typically 2 m), high-speed wireless links. The MICS band overcomes the limitations of dated inductive systems and facilitates the development of next-generation medical devices supporting improved patient healthcare. This is especially important, as escalating medical costs drive the growth of remote health monitoring.

Because of the signal propagation characteristics in the human body, compatibility with the incumbent users of the band (meteorological aids such as weather balloons), and its international availability, the 402–405 MHz band is well suited for such remote monitoring.

Transceiver Design Considerations

The design of transceivers for medical devices is challenged by the following basic requirements:

  • Low power consumption during 400 MHz communication is required. Implant battery power is limited, and the impedance of implant batteries is relatively high. This combination limits peak currents that may be drained from the supply. During communication sessions, current should be limited to <6 mA for most implantable devices.
  • The transceiver must operate in a low-power sleep mode, with the capability to look periodically for a wake-up signal.
  • Minimum external component count and small physical size are important factors. An RF module for a pacemaker must be no larger than ~5 x 5 x 10 mm. Furthermore, implant-grade components are expensive, and using high levels of integration may significantly reduce costs. Integration has the additional benefit of increasing overall system reliability.
  • Reasonable data rates are demanded; pacemaker applications are currently demanding >20 Kb/sec, with higher data rates projected for the future.
  • High reliability in both data transmission and system operation.
  • An operating range is typically >2 m because the MICS band is designed to improve upon the very-short-range inductive link. Longer operating ranges imply that good sensitivity is needed, because small antennas and body loss affect link budget and allowable range. Antenna matching and body loss can typically be more than 40 dB.
  • Selectivity is required and interference must be rejected.

The MICS regulations provide additional requirements. This section discusses some of the key requirements.

The MICS regulations require a system to perform a clear-channel assessment (CCA) in which the user scans all 10 of the 300-kHz channels and is allowed to transmit on the channel with the lowest ambient signal level (the least-noisy channel). The user can also choose to transmit on the first available channel with an ambient power below a certain threshold (as defined in the standard). The MICS standard requires that the external programmer carry out the scanning process. For this reason, the IMD transceiver should support a low-power method of sniffing for the presence of an external programmer signal.

MICS regulations provide an exception to the CCA procedure in the event of an emergency medical event. For clinically significant medical emergencies, the IMD may transmit immediately on any channel. For example, if an implanted ECG monitor or pacemaker detects a cardiac arrest, the device could transmit immediately to a monitoring base station that, in turn, calls an emergency response service.

Given the important requirements defined above, it is essential that medical device designers and system architects meet the demands of RF medical implant communication.

Figure 1. An implanted pacemaker uses the MICS band to communicate patient health and operating data to a base station, which transmits information to a physician's remote office.
(click to enlarge)

An example of a transceiver IC specifically designed for implanted medical devices operating in the 402–405 MHz MICS band is shown in Figure 1. The concepts of duty cycling, ultra-low-power circuit design, and high integration levels are incorporated, with specific attention paid to the special needs of IMD systems. This example will be used to highlight the concepts introduced in the following sections.

Achieving Low Power Consumption

Medical devices can be divided into two categories: those that use an internal nonrechargeable battery (e.g., pacemakers) and those that couple power inductively (e.g., cochlear implants). The former heavily duty-cycle the operation of systems to conserve power. The transceiver is off most of the time. Therefore, the off-state current and the current required to periodically look for a communicating device must be extremely low (<1 µA). In both cases, low power (<6 mA) is also required for both transmit and receive.

The primary philosophy for saving power in IMD electronics (RF or otherwise) is the simple concept of duty-cycling: operate systems for a short time period and with minimum current. Furthermore, low leakage currents must be ensured when systems are disabled.

By operating systems quickly, fixed power-consuming overheads (such as support circuitry, synthesizers, clocks, biasing, and regulators) have less time to drain the battery. There is always a limit to applying this principle, because circuit complexity will rise, and total energy requirements will increase if the operating time is too short. Therefore, an optimal time period usually exists where the total energy consumed is minimized.

For RF communication systems in medical implants, duty-cycling may be applied to all phases of operation. The IMD transceiver sniffing and start-up should be duty-cycled. In addition, duty-cycling and short on-times may be exploited during the actual 400 MHz communication sessions.

Duty-Cycling Normal Transmission. For minimum overall power consumption, defined in terms of joules per bit, it is recommended that implantable transceivers use the highest possible data rate that satisfies the application receiver-sensitivity requirements. Systems that require low data rates (even in the low kilohertz range) should buffer data, operate at the highest data rate possible, and exploit duty-cycling of the power states to reduce the average current consumption. Sending data in short bursts conserves power and reduces the potential time window for interference. In addition, the power supply decoupling requirements are more forgiving in systems with high battery impedance.

Figure 2. Highly integrated ultra-low-power MICS transceiver.
(click to enlarge)

The transceiver architecture presented in Figure 2 allows a user to select from a wide range of relatively high data rates (up to 800 Kb/sec) with varying receiver sensitivity. To facilitate this flexibility, the system uses frequency-shift-keyed (FSK) modulation with varying frequency deviations.

In low-power architectures, the modulation scheme should provide high data rates while supporting a simple radio architecture that meets the current consumption requirements. Quadrature amplitude modulation and Nyquist-filtered M-ary (or multiple) phase modulation both offer good bandwidth efficiencies. However, constant-envelope signals (i.e., FSK) are advantageous because they result in relaxed requirements on the linearity of the system. Of the available modulation schemes, FSK modulation has been found to provide a good compromise between data rate, complexity, and requirements on linearity.

FSK allows for a high-data-rate, low-power receiver. Note that more-complex modulation schemes are often seen in short-range RF communication standards, such as Bluetooth. These modulation schemes are not easily amenable to the truly ultra-low-power communication demanded by medical implants.

The design of the communication control should facilitate duty-cycling of the RF blocks when data are not ready to be transmitted. The transceiver in Figure 2 includes an integrated media access controller (MAC) that may be used to control the power state of analog circuitry when data communication is not immediately required or is being buffered. The MAC relieves the user of communication link control activities to the extent that the RF link is simply a memory-mapped peripheral.

Duty-Cycling Transceiver Wake-Up. Most implant applications use the MICS RF link infrequently because of their overriding need to conserve battery power. In very-low-power applications, the transceiver spends most of the time asleep in a very-low-current state and periodically sniffs for a wake-up signal.

This sniffing operation should be frequent enough to provide reasonable start-up latency and, because it will occur regularly, it should consume a very low current. It should also be immune to noise sources that invoke an erroneous start-up. In this situation, an on-off keyed (OOK) modulation scheme is recommended because the OOK scheme removes the need for a local oscillator and synthesizer in the receiver, both of which require time and power to start up.

The wake-up system shown in Figure 2 uses an ultra-low-power RF receiver to read OOK transmitted data. The receiver's main function is to detect the incoming signal from the programmer and then to activate the rest of the chip. The example shown may also be started directly by pin control, which allows either an external programmer to initiate communication or the implant itself to send an emergency communication.

Achieving High Data Integrity

The MICS Band

An ultra-low-power RF transmitter is a key component of the swallowable camera.
(click to enlarge)

For typical RF communications, given the plethora of noise and interference sources that exist, a reasonable design guideline is to assume a minimum raw radio bit-error rate of about 1 x 10 –3 errors/bit. The MAC shown in Figure 2 includes forward error correction (FEC) using a Reed-Solomon block code and CRC error detection. The design of appropriate FEC codes is interesting, with trade-offs existing between power consumed by retransmission of packets and power consumed by additional coding capability. FEC codes also provide an overall trade-off between final bit-error rate and system power consumption.

For a 30% overhead in bits used for error correction and detection, for example, an improvement can be obtained in bit-error rate of many orders of magnitude from 1 x 10–3 to at least 1 x 10–4 errors/bit. This improvement equates to an average time between errors, given a data rate of 200 Kb/sec, of four years under the impractical scenario of continuous operation. As noted earlier, in order to save battery power, the RF link in most applications will be used rarely.

Reed-Solomon codes are especially good at correcting burst errors. In addition, high data rates offer the opportunity to minimize the probability of interference in burst-noise cases frequently observed in an RF environment.

An ultra-low-power RF transmitter is used in this swallowable capsule. The transmitter sends images of the gastrointestinal tract.
(click to enlarge)

From an analog RF standpoint, immunity from interference places constraints on channel and band filtering and on low-noise amplifier (LNA) linearity because of intermodulation. Immunity also limits synthesizer phase noise caused by reciprocal mixing. Such constraints are often in direct conflict with low-power design and require skilled RF analog IC design expertise.

In some systems, an external standing-acoustic-wave (SAW) filter may be appropriate. This is more easily accommodated in the external programmer, because it is usually subject to worse interference scenarios than the IMD. A SAW filter also does not suffer as severely from space constraints. The external programmer is more susceptible to interference due to two factors: the attenuation of an interferer through the body before reaching the implant, and the normally smaller available power received by the external programmer from the implant.

The MICS standard allows for a maximum transmitted power (in air) of 25 mW. The external programmer may easily radiate at this power level; however, this will not be the case for an implant due to large antenna losses, body losses, and limitations in available implant transmit power. Good power amplifier design is important for maximizing the available implant transmit power. Nevertheless, battery performance and current limitations in the implant are the ultimate limiting factors.

Conclusion

The design considerations for implantable RF communication systems require careful consideration of transceiver design, power consumption, and data integrity. Moreover, when designing an implantable RF communication system, it is essential to consider the issues related to remote management of these devices via the MICS band. The transceiver IC discussed in this article uses an architecture that provides the basic concepts of achieving low power and high data integrity.

These concepts will lead the way to high-performance, low-power RF communication. Such transceiver technology will facilitate clinically significant improvements in healthcare for the development of next generation of medical implants.

Peter Bradley, PhD, is a project engineer and system architect with Zarlink Semiconductor's (Ottawa, ON, Canada) ultra-low-power communications division. He is responsible for the development of medical integrated circuits. Bradley holds a bachelor of engineering and master of biomedical engineering from the University of New South Wales, Australia, and a PhD in medical physics from the University of Wollongong, Australia. He can be reached at peter.bradley@zarlink.com.

Copyright ©2005 Medical Electronics Manufacturing

No votes yet